Midterms 2026, one week out: markets give Democrats about 91% for the House and 64% for the Senate
With the 3 November midterms a week away, Kalshi and Polymarket give Democrats a 90–92% chance of winning the House and 63–65% for the Senate, up from about 50% at the end of August. The odds for both chambers, the Senate races that decide control, how the prices moved and what to watch on election night.
In this article
A week before the midterms, prediction markets expect Democrats to win the House and are leaning their way in the Senate. Kalshi gives Democrats a 90% chance of winning the House and Polymarket 91.5%. For the Senate, Kalshi has Democrats at 63% and Polymarket at 64.5%. Election day is Tuesday, 3 November.
Prices on the morning of 26 October 2026.
The odds
| Question | Kalshi | Polymarket |
|---|---|---|
| House: Democrats win | 90% | 91.5% |
| House: Republicans win | 10% | 8.5% |
| Senate: Democrats win | 63% | 64.5% |
| Senate: Republicans win | 37% | 35.5% |
The combined balance-of-power market says the same thing. The single most likely result is a Democratic sweep of both chambers at 64% on both platforms, followed by a split Congress with a Democratic House and a Republican Senate at 26–26.5%. A Republican sweep is priced at 7.5% on Polymarket and about 10–11% on Kalshi, and a Democratic Senate with a Republican House at under 1%.
| Balance of power | Kalshi | Polymarket |
|---|---|---|
| Democrats sweep | 64% | 64.5% |
| Democratic House, Republican Senate | 26% | 26.5% |
| Republicans sweep | 10–11% | 7.5% |
| Democratic Senate, Republican House | 0.6% | 0.85% |
Polymarket's balance-of-power market alone has traded $37.9 million, the House market $23.5 million and the Senate market $9.2 million. Across all midterm markets, more than $1.4 billion had been traded by 23 October (the midterm trading numbers). Live prices: House on Polymarket · on Kalshi · Senate on Polymarket · on Kalshi.
How the odds moved
Source: Polymarket price history, 31 August to 26 October 2026
Show the dataHide the data
| House | Senate | |
|---|---|---|
| 31 Aug | 89% | 51% |
| 14 Sep | 87% | 53% |
| 28 Sep | 93% | 63% |
| 5 Oct | 93% | 67% |
| 12 Oct | 91% | 64% |
| 19 Oct | 91% | 65% |
| 23 Oct | 93% | 68% |
| 26 Oct | 92% | 65% |
The House has been priced as a near-certainty for Democrats since the summer, between about 86% and 93%. The Senate is where the market moved: at the end of August it was a coin flip, by late September Democrats were at about 62%, and they have stayed between 63% and 68% since, peaking at 67.5% on 23 October.
The Senate math
Republicans hold the Senate 53–47. With Vice President JD Vance breaking ties for Republicans, Democrats need 51 seats, a net gain of four. The seat-count market is slightly more bullish than the control market: it puts the chance of Republicans ending with 49 seats or fewer at about 67.5%, with "47 or fewer" the most likely single outcome at 40.5%.
That is because the races markets favour Democrats in add up to more than four flips:
| Race | Seat held by | Market favourite | Polymarket | Kalshi |
|---|---|---|---|---|
| North Carolina | Republicans | Roy Cooper (D) | 96% | 96.5% |
| Alaska | Republicans | Mary Peltola (D) | 74.5% | 75% |
| Ohio (special) | Republicans | Sherrod Brown (D) | 68.5% | 67% |
| Texas | Republicans | James Talarico (D) | 64.5% | 65% |
| Maine | Republicans | Troy Jackson (D) | 60% | 60% |
| Iowa | Republicans | Ashley Hinson (R) | 57.5% | 59% |
| Kansas | Republicans | Roger Marshall (R) | 62.5% | 61% |
| Nebraska | Republicans | Pete Ricketts (R) | 72.5% | 74% |
| Michigan | Democrats | Abdul El-Sayed (D) | 67.5% | 67% |
| Minnesota | Democrats | Peggy Flanagan (D) | 85.5% | 85% |
| New Hampshire | Democrats | Chris Pappas (D) | 92% | 90% |
| Georgia | Democrats | Jon Ossoff (D) | 95.5% | 94.5% |
Five Republican-held seats have a Democratic favourite. If those favourites all won and Democrats held everything they have, they would end with 52 seats. But three of the five (Ohio, Texas and Maine) are priced between 60% and 70%, so the chance of Democrats winning at least four of them is well short of certain, which is why Senate control sits in the mid-60s. Texas is the most-traded Senate race on either platform, at $6.5 million on Polymarket alone (Texas Senate odds).
Iowa and Kansas are the Republican seats Democrats could pick up beyond the five. Michigan is the Democratic seat most at risk, with Mike Rogers at about a third.
How much to trust a 64%
A 64% favourite should lose about one time in three. Election markets have a good record: in our study of 358,236 resolved markets, elections were among the easiest questions for markets to forecast, and favourites priced around 70¢ a day out won about 75% of the time. But a favourite is not a winner. On Sunday, Polymarket gave Flávio Bolsonaro about 79% on the morning of Brazil's runoff, and he lost (what happened in Brazil).
What to watch on election night
The first polls close at 6 p.m. Eastern in parts of Indiana and Kentucky, and the Senate picture usually starts to form after 7 p.m., when Georgia and New Hampshire close, and 7:30 p.m., when North Carolina and Ohio close. Alaska is among the last to report. Our guide to when races get called has the full schedule, and the 2026 election hub follows every Senate, House and governor market live.
Markets settle on the official outcome under each market's rules, not on media projections. Polymarket's House market, for example, counts runoffs and, if no party wins a majority of seats, goes by the party of the first Speaker elected, so prices can stay short of 0 and 100 for days in a close count.
