Each pollster counts once, with its latest poll. That poll is weighted by its age, its sample size and the pollster’s past accuracy, measured on the 2010-2022 elections. A poll commissioned by a party, a candidate or an allied group is corrected by the average gap between such polls and independent ones.
Map of the 35 races
How many seats for each side
Each bar gives the probability of a seat count, across 200,000 simulations of the election. Blue: outcomes where the Democrats and their allies hold the Senate. Red: outcomes where the Republicans keep it.
Seats by level of certainty
9 competitive races (under 85%)
34 D and 31 R seats not up
The races that will decide the Senate
The 5 races where each side keeps at least a one-in-four chance, from closest to least close. The expected margin is the centre of the forecast, in points; the bar gives each candidate’s chance of winning.
The races that decide the Senate
Of the 35 seats up, most are not competitive. Control comes down to a handful of races, mostly Republican-held seats.
Democrats need to hold these seats
and gain 4 of these to win a majority
Poll average: D +3.4 (30 polls, latest October 21)
Last election (2020): R +9.6
Poll average: D +0.8 (19 polls, latest October 20)
Last election (2020): R +8.6
Poll average: D +4.0 (28 polls, latest October 22)
Last election (2020): D +1.7
Poll average: D +4.6 (13 polls, latest October 21)
Last election (2020): D +5.2
Poll average: R +0.7 (17 polls, latest October 22)
Last election (2020): R +6.6
Every race, from most Democratic to most Republican
The chance shown is the favourite’s. The expected margin is the centre of the forecast, with the range that holds the result 9 times out of 10. The poll average weights each poll by its age, its sample and the pollster’s past accuracy. “Flip” marks a seat whose favourite is not the party that won it last time.
| State | Chance of winning | Expected margin |
|---|---|---|
| MassachusettsRegular seatEdward J. Markey / John Deaton | >99%D chance | D +31.9D +20 to D +44 |
| DelawareRegular seatChris Coons / Michael "Dr. Mike" Katz | >99%D chance | D +35.8D +22 to D +50 |
| ColoradoRegular seatJohn Hickenlooper / Mark Baisley | >99%D chance | D +29.0D +17 to D +41 |
| Rhode IslandRegular seatJohn F. Reed / Raymond T. McKay | >99%D chance | D +28.0D +16 to D +40 |
| New JerseyRegular seatCory Booker / Justin Murphy | >99%D chance | D +32.5D +18 to D +47 |
| OregonRegular seatJeff Merkley / David Brock Smith | >99%D chance | D +32.2D +18 to D +46 |
| VirginiaRegular seatMark R. Warner / Bert Mizusawa | >99%D chance | D +24.1D +12 to D +36 |
| New MexicoRegular seatBen R. Luján / Larry E. Marker (write-in primaire R) | >99%D chance | D +19.2D +7 to D +31 |
| IllinoisRegular seatJuliana Stratton / Don Tracy | 96%D chance | D +15.4D +1 to D +30 |
| GeorgiaRegular seatJon Ossoff / Mike Collins | 91%D chance | D +10.0R +2 to D +22 |
| North Carolina flipRegular seatRoy Cooper / Michael Whatley | 85%D chance | D +7.6R +4 to D +20 |
| New HampshireRegular seatChris Pappas / John E. Sununu | 77%D chance | D +5.3R +7 to D +17 |
| MinnesotaRegular seatPeggy Flanagan / Michele Tafoya | 73%D chance | D +4.6R +7 to D +17 |
| MichiganRegular seatAbdul El-Sayed / Mike Rogers | 60%D chance | D +1.8R +10 to D +14 |
| TexasRegular seatJames Talarico / Ken Paxton | 55%R chance | R +0.9R +13 to D +11 |
| MaineRegular seatTroy D. Jackson / Susan M. Collins | 57%R chance | R +1.3R +13 to D +11 |
| IowaRegular seatJosh Turek / Ashley Hinson | 74%R chance | R +4.6R +17 to D +7 |
| NebraskaRegular seatDan Osborn / Pete Ricketts | 76%R chance | R +8.3R +28 to I +12 |
| OhioSpecial electionSherrod Brown / Jon Husted | 76%R chance | R +5.1R +17 to D +7 |
| AlaskaRegular seatMary Peltola / Dan S. Sullivan | 82%R chance | R +6.6R +19 to D +5 |
| South CarolinaRegular seatAnnie Andrews / Darline Graham | 86%R chance | R +7.8R +20 to D +4 |
| KansasRegular seatAdam Hamilton / Roger Marshall | 86%R chance | R +7.8R +20 to D +4 |
| MontanaRegular seatSeth Bodnar / Kurt Alme | 87%R chance | R +13.8R +34 to I +6 |
| LouisianaRegular seat"Jamie" Davis / Julia Letlow | 89%R chance | R +8.9R +21 to D +3 |
| IdahoRegular seatTodd Achilles / Jim Risch | 94%R chance | R +18.9R +39 to I +1 |
| FloridaSpecial electionAngie Nixon / Ashley Moody | 95%R chance | R +11.9R +24 to 0 |
| South DakotaRegular seatBrian Bengs / Mike Rounds | 96%R chance | R +21.0R +41 to R +1 |
| MississippiRegular seatScott Colom / Cindy Hyde-Smith | 97%R chance | R +13.3R +25 to R +1 |
| ArkansasRegular seatHallie Shoffner / Tom Cotton | >99%R chance | R +19.6R +32 to R +8 |
| KentuckyRegular seatCharles Booker / Andy Barr | >99%R chance | R +20.2R +32 to R +8 |
| AlabamaRegular seatEverett Wess / Barry Moore | >99%R chance | R +22.3R +34 to R +10 |
| OklahomaRegular seatN'Kiyla Jasmine Thomas / Kevin Hern | >99%R chance | R +30.1R +42 to R +18 |
| TennesseeRegular seatMarquita Bradshaw / Bill Hagerty | >99%R chance | R +30.5R +42 to R +19 |
| WyomingRegular seatJames Byrd / Harriet Hageman | >99%R chance | R +44.8R +59 to R +31 |
| West VirginiaRegular seatRachel Fetty Anderson / Shelley Moore Capito | >99%R chance | R +53.4R +68 to R +39 |
Each seat’s last election, compared with 2026
The Senate is renewed by classes, so the reference is not a single national election. Left: each seat’s last result (2020 for regular seats, 2022 for the two special elections). Right: the expected margin for 2026. Drag the slider to move from one to the other. The four races where an independent faces a Republican have no point of comparison and stay beige.
Historical source: the seat's last archived Senate result, with its year. Differences are computed only when both axes are D−R.
How each figure is obtained
Three blocks, weighted by the distance to election day: the polls, the state’s history and the structure. Then the uncertainty, which gives the probability.
Alabama
Everett Wess (D) vs Barry Moore (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Cygnal | Oct. 17 | 500 | 35–50 |
| The Alabama Poll | Oct. 16 | 600 | 35–56 |
| RMG Research | Sept. 25 | 800 | 33–54 |
From expected margin to probability
- 90% range
- R +34.2 to R +10.3
- D chance (direct calculation)
- 0.1%
- D chance (200,000 simulations)
- 0.1%
Alaska
Mary Peltola (D) vs Dan S. Sullivan (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Data for Progress | Oct. 16 | 694 | 47–44 |
| Public Policy Polling | Oct. 15 | 578 | 48–47 |
| InsiderAdvantage | Oct. 10 | 1200 | 50–48 |
| Alaska Survey Research | Oct. 3 | 1512 | 51–49 |
| New York Times/Siena University | Oct. 1 | 504 | 50–43 |
| Quantus Insights | Sept. 30 | 758 | 46–46 |
| Trafalgar Group | Sept. 24 | 1079 | 46–47 |
| co/efficient | Sept. 17 | 799 | 46–48 |
2 older polls.
From expected margin to probability
- 90% range
- R +18.6 to D +5.4
- D chance (direct calculation)
- 18.2%
- D chance (200,000 simulations)
- 18.2%
Highest third candidate: Dan J. Sullivan (R), 3.5% on average. Their chance of winning is bounded by the polls and remains negligible.
Arkansas
Hallie Shoffner (D) vs Tom Cotton (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Hendrix College | Oct. 13 | 709 | 41–50 |
| GrayHouse | Oct. 4 | 550 | 39–50 |
| J.L. Partners | Aug. 18 | 803 | 35–51 |
From expected margin to probability
- 90% range
- R +31.6 to R +7.6
- D chance (direct calculation)
- 0.4%
- D chance (200,000 simulations)
- 0.4%
Colorado
John Hickenlooper (D) vs Mark Baisley (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- D +17.1 to D +41.0
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
Delaware
Chris Coons (D) vs Michael "Dr. Mike" Katz (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- D +21.6 to D +50.0
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
Florida · special election
Angie Nixon (D) vs Ashley Moody (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| co/efficient | Oct. 20 | 1014 | 40–48 |
| University of North Florida | Oct. 19 | 1471 | 45–49 |
| ActiVote | Oct. 15 | 400 | 47–53 |
| Univision/YouGov | Oct. 15 | 825 | 42–51 |
| MDW Communications | Oct. 10 | 2351 | 42–48 |
| Stetson University | Oct. 9 | 716 | 41–52 |
| St. Pete Polls | Oct. 7 | 836 | 45–47 |
| YouGov | Oct. 5 | 3813 | 44–49 |
5 older polls.
From expected margin to probability
- 90% range
- R +23.8 to D +0.1
- D chance (direct calculation)
- 5.2%
- D chance (200,000 simulations)
- 5.2%
Georgia
Jon Ossoff (D) vs Mike Collins (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| ActiVote | Oct. 20 | 400 | 54–46 |
| Quantus Insights | Oct. 17 | 755 | 52–45 |
| University of Georgia School of Public and International Affairs | Oct. 15 | 1000 | 48–40 |
| Trafalgar Group | Oct. 9 | 1101 | 51–43 |
| InsiderAdvantage | Sept. 23 | 1200 | 50–42 |
| Big Data Poll | Sept. 23 | 678 | 52–41 |
| YouGov | Sept. 18 | 3301 | 52–42 |
| Rasmussen Reports | Sept. 14 | 1019 | 51–42 |
From expected margin to probability
- 90% range
- R +2.0 to D +21.9
- D chance (direct calculation)
- 91.5%
- D chance (200,000 simulations)
- 91.5%
Idaho
Todd Achilles (I) vs Jim Risch (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | I–R |
|---|---|---|---|
| Peak Insights | Sept. 30 | 500 | 24–49 |
| Advanced Targeting Research | Sept. 16 | 700 | 50–39 |
| The Bullfinch Group | Aug. 7 | 608 | 38–33 |
| Change Research | July 30 | 1213 | 21–45 |
From expected margin to probability
- 90% range
- R +38.7 to I +0.9
- I chance (direct calculation)
- 5.8%
- I chance (200,000 simulations)
- 5.8%
Highest third candidate: Other, 11.0% on average. Their chance of winning is bounded by the polls and remains negligible.
Illinois
Juliana Stratton (D) vs Don Tracy (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| ActiVote | Oct. 20 | 400 | 58–42 |
| Public Policy Polling | Oct. 1 | 569 | 49–36 |
From expected margin to probability
- 90% range
- D +1.3 to D +29.6
- D chance (direct calculation)
- 96.3%
- D chance (200,000 simulations)
- 96.3%
Iowa
Josh Turek (D) vs Ashley Hinson (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| InsiderAdvantage | Oct. 22 | 1200 | 45–47 |
| co/efficient | Oct. 19 | 863 | 45–48 |
| Trafalgar Group | Oct. 15 | 1094 | 44–46 |
| GS Strategy Group/ New River Strategies | Oct. 13 | 602 | 46–44 |
| YouGov | Oct. 12 | 2245 | 47–46 |
| NPA Polling | Oct. 12 | 678 | 44–47 |
| Marist University | Oct. 11 | — | 48–45 |
| Emerson College | Oct. 6 | 650 | 45–47 |
9 older polls.
From expected margin to probability
- 90% range
- R +16.6 to D +7.3
- D chance (direct calculation)
- 26.2%
- D chance (200,000 simulations)
- 26.1%
Kansas
Adam Hamilton (D) vs Roger Marshall (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| InsiderAdvantage | Oct. 14 | 1200 | 47–48 |
| GBAO | Oct. 4 | 600 | 46–41 |
| YouGov | Oct. 4 | 2255 | 48–45 |
| Trafalgar Group | Oct. 1 | 1095 | 44–44 |
| New York Times/Siena University | Sept. 30 | 605 | 45–45 |
| Global Strategy Group | Sept. 16 | 800 | 49–47 |
| Emerson College | Sept. 16 | 750 | 45–43 |
| co/efficient | Sept. 10 | 915 | 44–49 |
1 older poll.
From expected margin to probability
- 90% range
- R +19.8 to D +4.1
- D chance (direct calculation)
- 14.0%
- D chance (200,000 simulations)
- 14.1%
Kentucky
Charles Booker (D) vs Andy Barr (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Global Strategy Group | Aug. 27 | 600 | 40–49 |
From expected margin to probability
- 90% range
- R +32.2 to R +8.2
- D chance (direct calculation)
- 0.3%
- D chance (200,000 simulations)
- 0.3%
Louisiana
"Jamie" Davis (D) vs Julia Letlow (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Hart Research Associates | Sept. 19 | 500 | 44–48 |
From expected margin to probability
- 90% range
- R +20.8 to D +3.1
- D chance (direct calculation)
- 11.2%
- D chance (200,000 simulations)
- 11.1%
Maine
Troy D. Jackson (D) vs Susan M. Collins (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Beacon Research/ Shaw & Co. Research | Oct. 20 | 1514 | 49–49 |
| InsiderAdvantage | Oct. 14 | 1200 | 49–47 |
| Suffolk University | Oct. 11 | 500 | 49–47 |
| Rasmussen Reports | Oct. 8 | 804 | 49–47 |
| The Washington Post/SSPG | Oct. 5 | 800 | 50–46 |
| Marist University | Oct. 4 | 1505 | 51–47 |
| CBS News/YouGov | Oct. 1 | 1136 | 50–50 |
| Trafalgar Group | Oct. 1 | 1091 | 46–46 |
11 older polls.
From expected margin to probability
- 90% range
- R +13.2 to D +10.7
- D chance (direct calculation)
- 43.1%
- D chance (200,000 simulations)
- 43.0%
Highest third candidate: Other, 2.1% on average. Their chance of winning is bounded by the polls and remains negligible.
Massachusetts
Edward J. Markey (D) vs John Deaton (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Suffolk University | Oct. 18 | 500 | 60–32 |
| UMass Amherst/YouGov | Oct. 12 | 800 | 54–28 |
| MassINC Polling Group | Sept. 22 | 800 | 52–26 |
| University of New Hampshire | Sept. 21 | 564 | 53–30 |
From expected margin to probability
- 90% range
- D +19.9 to D +43.9
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
Michigan
Abdul El-Sayed (D) vs Mike Rogers (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| InsiderAdvantage | Oct. 22 | 1200 | 47–46 |
| Marist University | Oct. 18 | — | 50–48 |
| Rasmussen Reports | Oct. 14 | 798 | 46–44 |
| The Tyson Group | Oct. 8 | 600 | 44–46 |
| Quantus Insights | Oct. 6 | 762 | 48–47 |
| YouGov | Oct. 4 | 3949 | 49–48 |
| Mitchell Research & Communications | Oct. 1 | 682 | 46–41 |
| Z to A Research | Oct. 1 | 483 | 49–42 |
20 older polls.
From expected margin to probability
- 90% range
- R +10.2 to D +13.8
- D chance (direct calculation)
- 59.7%
- D chance (200,000 simulations)
- 59.9%
Minnesota
Peggy Flanagan (D) vs Michele Tafoya (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Rasmussen Reports | Oct. 21 | 1182 | 48–45 |
| Trafalgar Group | Oct. 15 | 1079 | 48–45 |
| KSTP/SurveyUSA | Oct. 12 | 718 | 45–42 |
| Mason-Dixon Polling & Strategy | Oct. 7 | 800 | 47–43 |
| Morning Consult | Oct. 7 | 1000 | 45–41 |
| GQR | Oct. 4 | 800 | 50–42 |
| Big Data Poll | Sept. 30 | 622 | 40–38 |
| InsiderAdvantage | Sept. 29 | 1200 | 46–45 |
5 older polls.
From expected margin to probability
- 90% range
- R +7.4 to D +16.5
- D chance (direct calculation)
- 73.5%
- D chance (200,000 simulations)
- 73.4%
Mississippi
Scott Colom (D) vs Cindy Hyde-Smith (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Change Research | Oct. 22 | 842 | 40–43 |
| Data for Progress | Aug. 15 | 1018 | 38–44 |
From expected margin to probability
- 90% range
- R +25.2 to R +1.3
- D chance (direct calculation)
- 3.4%
- D chance (200,000 simulations)
- 3.5%
Montana
Seth Bodnar (I) vs Kurt Alme (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | I–R |
|---|---|---|---|
| Quantus Insights | Oct. 18 | 667 | 28–44 |
| Peak Insights | Oct. 12 | 500 | 29–44 |
| Concord Public Opinion Partners | Oct. 8 | 405 | 37–35 |
| American Pulse Research | Sept. 24 | 604 | 26–45 |
| Aspect Strategic | Sept. 15 | 715 | 32–44 |
| Rutgers University Eagleton Institute of Politics | Sept. 7 | 845 | 30–38 |
| Hart Research | Aug. 4 | 800 | 30–45 |
| Tavern Research | July 27 | 517 | 25–49 |
1 older poll.
From expected margin to probability
- 90% range
- R +33.6 to I +6.1
- I chance (direct calculation)
- 12.7%
- I chance (200,000 simulations)
- 12.6%
Highest third candidate: Alani Bankhead (D), 15.8% on average. Their chance of winning is bounded by the polls and remains negligible.
Nebraska
Dan Osborn (I) vs Pete Ricketts (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | I–R |
|---|---|---|---|
| GBAO | Oct. 22 | 800 | 46–45 |
| Impact Research | Oct. 20 | 600 | 47–49 |
| National Public Affairs | Oct. 18 | 598 | 43–46 |
| YouGov | Oct. 3 | 1809 | 48–47 |
| co/efficient | Sept. 30 | 1113 | 40–45 |
| SurveyUSA | Sept. 13 | 503 | 46–42 |
From expected margin to probability
- 90% range
- R +28.2 to I +11.5
- I chance (direct calculation)
- 24.5%
- I chance (200,000 simulations)
- 24.3%
Highest third candidate: Other, 3.7% on average. Their chance of winning is bounded by the polls and remains negligible.
New Hampshire
Chris Pappas (D) vs John E. Sununu (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Saint Anselm College | Oct. 20 | 1672 | 50–43 |
| National Public Affairs | Oct. 12 | 600 | 50–45 |
| YouGov | Oct. 5 | 1520 | 51–43 |
| Trafalgar Group | Sept. 27 | 1082 | 47–41 |
| New York Times/Siena University | Sept. 22 | 613 | 50–45 |
| University of New Hampshire | Sept. 21 | 1418 | 50–42 |
| InsiderAdvantage | Sept. 17 | 1200 | 48–40 |
| co/efficient | Sept. 11 | 958 | 46–46 |
2 older polls.
From expected margin to probability
- 90% range
- R +6.6 to D +17.3
- D chance (direct calculation)
- 76.8%
- D chance (200,000 simulations)
- 76.8%
New Jersey
Cory Booker (D) vs Justin Murphy (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- D +18.3 to D +46.6
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
New Mexico
Ben R. Luján (D) vs Larry E. Marker (write-in primaire R) (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Research & Polling Inc. | Oct. 9 | 1125 | 53–37 |
| SurveyUSA | Sept. 30 | 567 | 54–36 |
From expected margin to probability
- 90% range
- D +7.3 to D +31.2
- D chance (direct calculation)
- 99.6%
- D chance (200,000 simulations)
- 99.6%
North Carolina
Roy Cooper (D) vs Michael Whatley (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Harper Polling | Oct. 22 | 600 | 49–39 |
| ActiVote | Oct. 20 | 400 | 57–43 |
| SurveyUSA | Oct. 11 | 643 | 54–40 |
| Trafalgar Group | Oct. 10 | 1091 | 49–41 |
| Elon University/YouGov | Oct. 8 | 561 | 52–39 |
| Change Research | Oct. 7 | 829 | 50–40 |
| East Carolina University | Oct. 3 | 810 | 50–43 |
| Catawba College | Sept. 30 | 885 | 48–37 |
11 older polls.
From expected margin to probability
- 90% range
- R +4.3 to D +19.6
- D chance (direct calculation)
- 85.3%
- D chance (200,000 simulations)
- 85.2%
Ohio · special election
Sherrod Brown (D) vs Jon Husted (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| InsiderAdvantage | Oct. 22 | 1200 | 47–44 |
| Verasight/VoteHub | Oct. 20 | 531 | 45–41 |
| Beacon Research/ Shaw & Co. Research | Oct. 19 | 1592 | 52–47 |
| ActiVote | Oct. 18 | 400 | 50–50 |
| Marist University | Oct. 18 | — | 48–44 |
| Rasmussen Reports | Oct. 14 | 901 | 47–49 |
| GS Strategy Group/New River Strategies | Oct. 13 | 959 | 49–43 |
| Bowling Green State University/YouGov | Oct. 9 | 1000 | 48–42 |
10 older polls.
From expected margin to probability
- 90% range
- R +17.1 to D +6.9
- D chance (direct calculation)
- 24.2%
- D chance (200,000 simulations)
- 24.3%
Oklahoma
N'Kiyla Jasmine Thomas (D) vs Kevin Hern (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| CHS & Associates | Aug. 6 | 500 | 25–49 |
From expected margin to probability
- 90% range
- R +42.1 to R +18.1
- D chance (direct calculation)
- 0.0%
- D chance (200,000 simulations)
- 0.0%
Oregon
Jeff Merkley (D) vs David Brock Smith (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- D +18.0 to D +46.4
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
Rhode Island
John F. Reed (D) vs Raymond T. McKay (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Emerson College/WPRI-TV | Oct. 11 | 800 | 59–25 |
| Embold Research | Oct. 8 | 815 | 55–32 |
| University of New Hampshire | Sept. 21 | 595 | 53–33 |
From expected margin to probability
- 90% range
- D +16.0 to D +40.0
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
South Carolina
Annie Andrews (D) vs Darline Graham (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Impact Research | Oct. 21 | 800 | 45–44 |
| Trafalgar Group | Oct. 18 | 1095 | 46–49 |
| Rasmussen Reports | Oct. 8 | 837 | 45–49 |
| InsiderAdvantage | Sept. 29 | 1200 | 45–46 |
| Abacus Data | Aug. 28 | 306 | 42–55 |
From expected margin to probability
- 90% range
- R +19.8 to D +4.2
- D chance (direct calculation)
- 14.2%
- D chance (200,000 simulations)
- 14.2%
South Dakota
Brian Bengs (I) vs Mike Rounds (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | I–R |
|---|---|---|---|
| Emerson College | Oct. 20 | 803 | 42–50 |
| Quantus Insights | Oct. 19 | 633 | 37–57 |
| Public Opinon Strategies | Oct. 16 | 500 | 37–54 |
| Mason-Dixon Polling & Strategy | Oct. 14 | 500 | 37–56 |
| Concord Public Opinion Partners | Oct. 8 | 400 | 43–42 |
| Zenith Research | Sept. 27 | 500 | 38–44 |
| Public Policy Polling | Sept. 9 | 629 | 42–45 |
| Lake Research Partners | Aug. 30 | 500 | 51–44 |
1 older poll.
From expected margin to probability
- 90% range
- R +40.8 to R +1.1
- I chance (direct calculation)
- 4.1%
- I chance (200,000 simulations)
- 4.1%
Highest third candidate: Other, 5.3% on average. Their chance of winning is bounded by the polls and remains negligible.
Tennessee
Marquita Bradshaw (D) vs Bill Hagerty (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| Beacon Poll | Sept. 2 | 1200 | 31–51 |
| Targoz Market Research | Aug. 26 | 607 | 33–58 |
From expected margin to probability
- 90% range
- R +42.5 to R +18.5
- D chance (direct calculation)
- 0.0%
- D chance (200,000 simulations)
- 0.0%
Texas
James Talarico (D) vs Ken Paxton (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
| Pollster | Field end | Sample | D–R |
|---|---|---|---|
| ActiVote | Oct. 21 | 400 | 51–49 |
| Rasmussen Reports | Oct. 21 | 965 | 45–45 |
| Verasight/VoteHub | Oct. 20 | 509 | 46–41 |
| Beacon Research/ Shaw & Co. Research | Oct. 19 | 1398 | 49–49 |
| Emerson College | Oct. 19 | 1000 | 48–46 |
| Pollfinity Research | Oct. 18 | 278 | 48–43 |
| Texas Public Opinion Research | Oct. 18 | 969 | 51–47 |
| CBS News/YouGov | Oct. 16 | 1492 | 51–49 |
22 older polls.
From expected margin to probability
- 90% range
- R +12.9 to D +11.1
- D chance (direct calculation)
- 45.1%
- D chance (200,000 simulations)
- 45.1%
Virginia
Mark R. Warner (D) vs Bert Mizusawa (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- D +12.2 to D +36.1
- D chance (direct calculation)
- 100.0%
- D chance (200,000 simulations)
- 100.0%
West Virginia
Rachel Fetty Anderson (D) vs Shelley Moore Capito (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- R +67.5 to R +39.2
- D chance (direct calculation)
- 0.0%
- D chance (200,000 simulations)
- 0.0%
Wyoming
James Byrd (D) vs Harriet Hageman (R)What drives the forecast
Effect of each element on the expected margin, in points. The total is the forecast margin.
How the forecast has moved
Each candidate's chance of winning, out of 100 (the two add up to 100%). Curve since August 1, recomputed day by day with only the data published at each date.
Two-candidate vote share. Band: 90% range. Curve since August 1, recomputed day by day with only the data published at each date.
Polls used
No poll published for this race.
From expected margin to probability
- 90% range
- R +59.0 to R +30.6
- D chance (direct calculation)
- 0.0%
- D chance (200,000 simulations)
- 0.0%
How the forecast is built
The same method as Le Millénaire’s other forecasts (France, Germany, Brazil, Israel), adapted to the US Senate.
History starts from the state’s 2024 presidential margin and adds today’s generic ballot. It also gives an edge to a sitting senator on the ballot, appointed ones included, and a smaller edge to a former senator or governor of the state. Both effects were learned on 148 races from 2014 to 2022. Structure adds the seat’s last election and the candidates’ profiles.
The weight of the polls rises as the vote approaches. Three weeks out, they count for more than history and structure combined.
200,000 draws. A shared national shock links the Democrat versus Republican races; each race also keeps its own uncertainty.
What the method gave on past elections
Out-of-sample test on the 132 Senate races from 2014 to 2022: for each year, the model learns only from the others. Error in points of margin between the two candidates.
| Method | Mean error | Root mean square error | Winner called |
|---|---|---|---|
| Three blocks, Le Millénaire method (used) | 4.58 | 5.92 | 90.9% |
| Polls only, simple average | 4.94 | 6.33 | 90.9% |
| History only | 6.60 | 9.99 | 83.3% |
| Structure only | 7.52 | 10.85 | 82.6% |
- Each race is simulated as a two-way contest. A third candidate is bounded by the polls, without simulating their score.
- Polls come from secondary sources; checking against the pollster’s document is manual.
- Past errors do not guarantee accuracy in 2026. The national control probability has no historical test of its own.
Where the data come from
Wikipedia and 270toWin, collected daily, with the link to the pollster’s document when it is cited. Pollster reliability ratings: FiveThirtyEight public archive, 2010-2022 elections. Polls taken into account up to October 24, 2026.
2024 presidential vote by state: certified results published by the Federal Election Commission (FEC). Generic ballot: average of national polls (Wikipedia, PollsMax).
Officially registered candidates, checked race by race; last certified Senate results for each seat.
Party symbols. The donkey and the elephant are the symbols of the Democratic and Republican parties (Wikimedia Commons files). Used for identification; no affiliation.
Pollsters used (109 pollsters, number of polls)
- InsiderAdvantage13
- Quantus Insights12
- Trafalgar Group12
- Rasmussen Reports11
- YouGov11
- co/efficient10
- New York Times/Siena University8
- ActiVote6
- Beacon Research/ Shaw & Co. Research6
- Big Data Poll6
- Abacus Data5
- Emerson College5
- Fabrizio Ward/ Impact Research5
- Marist University5
- Suffolk University5
- Change Research4
- Public Policy Polling4
- SSRS4
- University of New Hampshire4
- GBAO3
- Global Strategy Group3
- Impact Research3
- SurveyUSA3
- CBS News/YouGov2
- Concord Public Opinion Partners2
- Cygnal2
- Data for Progress2
- GQR2
- GS Strategy Group/ New River Strategies2
- Guidant Polling & Strategy2
- Hart Research2
- Mason-Dixon2
- Mason-Dixon Polling & Strategy2
- National Public Affairs2
- Peak Insights2
- Targoz Market Research2
- The Washington Post/SSPG2
- TIPP Insights2
- Univision/YouGov2
- Verasight/VoteHub2
- Advanced Targeting Research1
- Alaska Survey Research1
- American Pulse Research1
- Aspect Strategic1
- Beacon Poll1
- Bowling Green State University/YouGov1
- Catawba College1
- Catawba College/YouGov1
- CHS & Associates1
- Cygnal/Beacon Research1
- Deep Root Analytics1
- East Carolina University1
- Elon University/YouGov1
- Embold Research1
- Emerson College/WPRI-TV1
- EPIC-MRA1
- Fabrizio, Lee, & Associates1
- Glengariff Group1
- GrayHouse1
- GS Strategy Group/New River Strategies1
- Harper Polling1
- Hart Research Associates1
- Hendrix College1
- High Point University1
- High Point University/YouGov1
- J.L. Partners1
- KSTP/SurveyUSA1
- KTSP/SurveyUSA1
- Lake Research Partners1
- Maine People's Resource Center1
- MassINC Polling Group1
- MDW Communications1
- Michigan State University/YouGov1
- Mitchell Research & Communications1
- Morning Consult1
- NPA Polling1
- Opinion Diagnostics1
- Overton Insights1
- Pollfinity Research1
- Public Opinon Strategies1
- Pulse Decision Science1
- ReconMR/Siena University1
- Research & Polling Inc.1
- RMG Research1
- Rutgers University Eagleton Institute of Politics1
- Saint Anselm College1
- St. Pete Polls1
- Stetson University1
- Stratus Intelligence1
- Susquehanna Polling & Research1
- Tavern Research1
- Texas A&M University/ReconMR/Siena University1
- Texas Public Opinion Research1
- Texas Southern University1
- The Alabama Poll1
- The Bullfinch Group1
- The Tyson Group1
- Torchlight Strategies1
- Tulchin Research1
- UMass Amherst/YouGov1
- UMass Lowell/YouGov1
- University of Georgia School of Public and International Affairs1
- University of Houston1
- University of Massachusetts/YouGov1
- University of North Florida1
- University of Texas/Texas Politics Project1
- YouGov Blue1
- Z to A Research1
- Zenith Research1
Links to the 229 poll documents
A forecast, not a certainty
Figures are recalculated every day with the published polls. Latest calculation: October 24, 2026.
A race given at 70% is lost by the favourite three times in ten. Joint control of the House and the Senate is on the overview page.
Run SENATE-COMMON-STATE-POLL-REFRESH-20261024 · 200,000 draws