2026 midterm election predictions: who will win the Senate, the House and the key races
Eleven days out, prediction markets and forecast models agree that Democrats are favourites to win the House and lean towards them in the Senate. Every forecast side by side, the seat-by-seat distribution, the Senate races and governorships that decide it, and how the predictions could be wrong, with charts.
In this article
- 2026 midterm predictions at a glance
- The Senate: the seat math
- The Senate races that decide it
- The House: how big a majority?
- Redistricting and the House battlefield
- Governors: the races to watch
- The most-traded midterm markets
- Markets vs models: why they differ
- How these predictions could be wrong
- When the key races should be called
- After the election: when the markets settle
- How to follow election night
- 2026 midterm predictions: frequently asked questions
- Keep reading
Democrats are favoured to win both chambers of Congress on 3 November 2026. Prediction markets give them about a 91–92.5% chance of winning the House and about 67% for the Senate. Forecast models agree on the direction but not the size: Nate Silver's Silver Bulletin puts Democrats near 75% for the Senate, while Decision Desk HQ and Poliwave have them at about 55–56%.
With the election eleven days away, here is every major prediction side by side, what the markets expect seat by seat, the races and governorships that decide it, and how these predictions could be wrong.
Prices as of 22 October 2026; model figures as published around 19–21 October. Live: our 2026 election hub shows Kalshi and Polymarket for every race.
2026 midterm predictions at a glance
| Forecast | Senate: Democrats | House: Democrats |
|---|---|---|
| Polymarket | 67.5% | 92.5% |
| Kalshi | 67% | 91.5% |
| Silver Bulletin (Nate Silver) | about 75% | — |
| Decision Desk HQ | 56% (51 seats projected) | — |
| Poliwave | 55% (51 seats median) | — |
Source: Polymarket and Kalshi, 22 October 2026; Silver Bulletin, Decision Desk HQ and Poliwave forecasts as reported 19–21 October 2026
Show the dataHide the data
| Value | |
|---|---|
| Silver Bulletin | 75.0% |
| Polymarket | 67.5% |
| Kalshi | 67.0% |
| Decision Desk HQ | 56.0% |
| Poliwave | 55.0% |
Every forecast has Democrats ahead, which is itself a big change: a year ago Polymarket gave them under 30% for the Senate. The disagreement is about how far ahead. The markets sit in the middle of the models.
The Senate: the seat math
Republicans hold the Senate 53–47. Because Vice President JD Vance breaks ties, Democrats need 51 seats, a net gain of four. Kalshi lists a contract for every possible number of Democratic seats after the election, which gives the clearest picture of what traders expect:
Source: Kalshi, “How many Senate seats will Democrats hold after the Midterms?” (both strike ranges), 22 October 2026; raw prices add up to about 109%
Show the dataHide the data
| Value | |
|---|---|
| ≤44 | 1.9% |
| 45 | 0.8% |
| 46 | 1.8% |
| 47 | 3.0% |
| 48 | 4.6% |
| 49 | 5.3% |
| 50 | 12.9% |
| 51 | 9.2% |
| 52 | 21.2% |
| 53 | 12.9% |
| 54 | 11.0% |
| 55 | 7.5% |
| 56 | 4.0% |
| 57 | 1.9% |
| 58+ | 1.8% |
The single most likely outcome is 52 Democratic seats (about 21%), and the distribution puts roughly 70% on 51 or more, in line with the 67% control markets. The 13% bump at 50 seats is the nightmare scenario for Democrats: a 50–50 Senate that Republicans would still control through the vice president's tie-breaking vote. Live: Democratic Senate seats on Kalshi.
The Senate races that decide it
Source: Polymarket Senate race markets, 22 October 2026
Show the dataHide the data
| Value | |
|---|---|
| North Carolina | 96.6% |
| New Hampshire | 90.5% |
| Minnesota | 87.5% |
| Alaska | 74.5% |
| Ohio (special) | 71.5% |
| Michigan | 69.5% |
| Texas | 67.5% |
| Maine | 65.5% |
| Iowa | 43.5% |
| Kansas | 38.5% |
| Nebraska | 28.5% |
If every favourite won, Democrats would flip North Carolina, Alaska, Ohio, Texas and Maine and hold all of their own seats, for 52, exactly the market's most likely outcome. The path runs through four Republican seats where Democrats are favoured but not safe:
- Texas: James Talarico against Ken Paxton, the most expensive race on Polymarket at about $5.5 million traded.
- Maine: Troy Jackson against Susan Collins, a five-term senator who has survived tough races before. Maine's ranked-choice count can take days.
- Ohio: Sherrod Brown, a former senator, against Jon Husted in a special election (Ohio Senate).
- Alaska: Mary Peltola against Dan Sullivan, with ranked-choice voting and results that can take a week or more.
On the other side, Iowa is the Republicans' best hold among the competitive seats, and Kansas and Nebraska are long shots for Democrats and an independent. Michigan is the only Democratic seat where Republicans have a real chance, at about 30%. Every race: Senate hub.
The House: how big a majority?
Democrats need 218 seats for a majority. Kalshi's seat markets show traders expect them to clear it comfortably:
Source: Kalshi, “How many House seats will Democrats hold after the Midterms?”, 22 October 2026; raw prices add up to about 107%
Show the dataHide the data
| Value | |
|---|---|
| <210 | 3.0% |
| 210–213 | 2.8% |
| 214–217 | 3.4% |
| 218–221 | 2.2% |
| 222–225 | 4.5% |
| 226–229 | 7.6% |
| 230–233 | 15.0% |
| 234–237 | 15.9% |
| 238–241 | 13.1% |
| 242–245 | 9.4% |
| 246–249 | 7.2% |
| 250+ | 15.9% |
The centre of the distribution is 230–241 seats, a majority of 25 to 47, and about 91% of the probability sits at 218 or more, matching the House control markets. The long tail at 250 or more (about 16%) is the "wave" scenario. Live: House control on Polymarket · on Kalshi · Democratic seats on Kalshi · House hub.
Redistricting and the House battlefield
This House fight is being fought partly on new maps. In 2025 Texas redrew its districts mid-decade to add Republican seats, and California voters responded in November 2025 by approving Proposition 50, which redrew their map to add Democratic seats. Kalshi's state seat markets show how traders expect those maps to play out:
| State | Seats | Most likely Democratic total (Kalshi) | Chance |
|---|---|---|---|
| California | 52 | 48 | 76–79% |
| Texas | 38 | 11, 12 or 13 | about 21–24% each |
| Virginia | 11 | 8 or 9 | about 34–35% each |
California's map is doing what it was drawn to do; in Texas, the spread of outcomes is wide, which says more about the national environment than about the map. Many individual districts have traded less than $100,000, so their prices are best read as rough guides.
Governors: the races to watch
Thirty-six governorships are on the ballot. Markets expect several to change party:
| State | Held by | Favourite (Polymarket) | Chance |
|---|---|---|---|
| Iowa | R | Rob Sand (D) | 88% |
| Alaska | R | Jonathan Kreiss-Tomkins (D) | 87.5% |
| Ohio | R | Amy Acton (D), against Vivek Ramaswamy | 79.5% |
| Nevada | R | Aaron Ford (D), against Governor Joe Lombardo | 66.5% |
| Georgia | R | Keisha Lance Bottoms (D) | 62.5% |
| Kansas | D | Ty Masterson (R) | 73.5% |
| Wisconsin | D | David Crowley (D) | 80.5% |
| Arizona | D | Governor Katie Hobbs (D) | 93.5% |
| California | D | Xavier Becerra (D) | 96.6% |
Kansas is the one Democratic governorship markets expect Republicans to take. California's race, at more than $42 million traded, is the biggest governor market of the cycle (California governor odds · Ohio governor odds · all governors).
The most-traded midterm markets
Source: Polymarket, 22 October 2026
Show the dataHide the data
| Value | |
|---|---|
| California governor | $42.2M |
| Balance of power | $29.7M |
| House control | $20.2M |
| Senate control | $7.8M |
| Texas Senate | $5.5M |
| Maine Senate | $2.4M |
| Michigan Senate | $2.1M |
| Alaska Senate | $1.6M |
| Ohio Senate | $1.5M |
Volume matters for how much weight a price deserves. The chamber and balance-of-power markets, with tens of millions traded, aggregate far more money and opinion than a Senate race with $1–2 million, let alone a House district with a few thousand dollars.
Markets vs models: why they differ
- Correlated polling error. Models such as Decision Desk HQ's build in the chance that polls miss in the same direction across many states, as they did in 2016 and 2020. That pulls their Senate numbers towards 50%.
- Different inputs. Each model weights polls and fundamentals its own way; Silver Bulletin's Senate number has risen steadily since August. Markets add money, momentum and information no model has, from early-vote data to campaign spending.
- Who trades. Midterm markets are big (more than $750 million traded on 2026 election markets, according to a CNN analysis), but concentrated: the top 1% of Polymarket wallets account for about 70% of congressional betting volume (the numbers).
Neither is reliably better. Markets beat polls far from election day in some cycles and missed badly in others, such as the 2022 Senate (how accurate are prediction markets · Brier score explained).
How these predictions could be wrong
- A polling miss. Most Democratic pickups are favourites by single digits in the polls. A two- or three-point error towards Republicans would put Maine, Texas, Ohio and Alaska back in play at once.
- Turnout. Midterms are decided by who shows up. A higher-than-expected Republican turnout, or a drop-off among younger voters, would shrink the margins.
- Late events. Eleven days is enough for a scandal, an economic shock or a foreign-policy crisis to move a close race.
- Slow counts. Maine and Alaska use ranked-choice voting; Georgia holds runoffs if no one wins a majority; California counts for weeks. Control of either chamber may not be known on election night, and markets resolve on the final result, not the first calls (how markets resolve).
- Thin markets. Many House districts have traded under $100,000, so their prices are easy to move (Kalshi's low-volume warnings).
When the key races should be called
Polls close at different times, and some states count far faster than others. Times are Eastern on 3 November:
| Race | Polls close (Eastern) | What to expect |
|---|---|---|
| North Carolina, Ohio | 7:30 p.m. | Both count early votes quickly; first strong signals of the night |
| Maine, New Hampshire | 8 p.m. (all polls closed) | Maine may need a ranked-choice tabulation, which takes days |
| Michigan | 8 p.m. (most of the state) | The one Democratic seat Republicans could take |
| Texas | 8 p.m. (most of the state) | Big early-vote batches come first |
| Kansas, Nebraska | 8 p.m. and 9 p.m. | Nebraska's independent race could be slow |
| Iowa | 9 p.m. | A Republican-held seat that tests the size of any wave |
| Alaska | Midnight (1 a.m. in the Aleutians) | Ranked-choice and late absentee ballots; often a week or more |
Across the country 35 Senate seats are on the ballot, including special elections. Control of the Senate could hinge on Maine or Alaska, so a final answer may take days.
After the election: when the markets settle
- Race markets usually resolve once major news organisations call the race or the result is certified, depending on each market's rules. Close races wait.
- Chamber control markets resolve on which party holds the majority when the new Congress meets on 3 January 2027, not on election-night projections.
- Kalshi's seat-count markets settle on the number of seats each party holds on 1 February 2027, which leaves time for recounts, runoffs and ranked-choice counts.
So money in a close race can stay tied up for weeks after 3 November (how Kalshi payouts work).
How to follow election night
Prices will move faster than any network call. Early results can mislead because some states count mail ballots first and others last; markets racing towards 95% are usually a better signal than raw vote counts. Our guide explains the timeline: how to follow live election odds. Everything about trading election markets, from legality to fees: election betting odds guide.
2026 midterm predictions: frequently asked questions
Who will win the Senate in 2026?
Democrats are favoured: about 67% on Kalshi and Polymarket, 75% in Silver Bulletin's model and 55–56% in Decision Desk HQ's and Poliwave's.
Who will win the House in 2026?
Democrats, with about a 91–92.5% chance in the markets; Kalshi's seat markets point to a majority of roughly 230–241 seats.
How many seats do Democrats need to win the Senate?
51, a net gain of four. At 50–50, Vice President JD Vance breaks ties for Republicans.
Which Senate seats are most likely to flip?
North Carolina, Alaska, Ohio, Texas and Maine, all held by Republicans, where markets favour Democrats.
When is the 2026 midterm election?
Tuesday, 3 November 2026.
How many Senate seats are up in 2026?
35, including special elections such as Ohio's.
Could control of Congress be unknown on election night?
Yes, particularly the Senate if it comes down to Maine or Alaska, where ranked-choice counts and late ballots can take days or weeks.
Are prediction markets more accurate than polls?
Often, especially far from election day, but not always. They work best read alongside polls and models.
