Prediction markets have made up their mind about the House and are still arguing about the Senate. As of September 21, 2026, six weeks before the November 3 midterms, Democrats trade at 91.5¢ to win the House on Polymarket and 91.3–91.5¢ on Kalshi. The Senate is closer: Democrats sit at 60.5¢ on Polymarket and 59–60¢ on Kalshi, a clear favourite but a long way from settled.
Those 2026 midterm election odds come from two separate order books with different users, different settlement currencies and different fee schedules, and they agree to within a cent. That agreement is information in itself. When the two venues disagree, somebody is thin or somebody is wrong; when they match this closely on markets this heavily traded, you are looking at about as clean a consensus as exists anywhere.
Below is every number that matters: chamber control, the combined balance-of-power market, the Senate races that decide the majority, how the race prices add up, and what to read before you put money on any of it.
| Market | Polymarket | Kalshi (bid / ask) | What it implies |
|---|---|---|---|
| House: Democrats | 91.5¢ | 91.3¢ / 91.5¢ | Roughly 91% |
| House: Republicans | 8.5¢ | 9.0¢ / 9.1¢ | Roughly 9% |
| Senate: Democrats | 60.5¢ | 59¢ / 60¢ | Roughly 60% |
| Senate: Republicans | 39.5¢ | 40¢ / 41¢ | Roughly 40% |
| Democratic sweep | 60.5¢ | — | Both chambers |
| Democratic House, Republican Senate | 30.5¢ | — | Split Congress |
| Republican sweep | 7.5¢ | — | Both chambers |
All prices were taken from each venue's public market data on the morning of September 21, 2026. For live numbers, our House control tracker and Senate control tracker compare both venues every ten minutes. A contract at 91.5¢ pays $1 if its outcome happens, so the price reads directly as a probability. If the jump from cents to odds is unfamiliar, our explainer on prediction market odds walks through it, and how prediction markets work covers the mechanics underneath.
House Control Odds: Democrats Near 91%
A 91¢ contract means the market thinks Republicans hold the House roughly one time in eleven. That is a strong position six weeks out, and history explains most of it. Since the Civil War, the president's party has gained House seats in a midterm only three times: 1934, 1998 and 2002. Every other midterm has gone the other way, often by a lot.
The arithmetic is also forgiving. Republicans won 220 seats in 2024, which means Democrats need a net gain of only three to take the chamber. The 2025 redistricting fight muddied individual districts — Texas redrew its map to add Republican-leaning seats and California answered with Proposition 50 — but the markets have had a year to digest both maps, and the price has kept climbing.
Here is what the number means in money. Buying Democratic House control at 91.5¢ returns 8.5¢ per contract if it lands, a 9.3% gain in about six weeks. Buying the Republican side at 8.5¢ returns $1 per contract, just under twelve times the stake, if the market is wrong. Neither is a free lunch. The first is a large bet to make a small return, and the second is priced about where the base rate says it should be.
Volume is heavy on both venues. Polymarket's House market has traded about $12 million. On Kalshi, roughly 12.3 million Democratic contracts and 21.6 million Republican contracts have changed hands, with open interest of about 22 million contracts across both sides. More Republican contracts trade because each one is cheaper — a dollar buys eleven Republican contracts and barely one Democratic one — so contract counts overstate how much money sits on the underdog. Our Kalshi review explains how the exchange reports volume.
Senate Control Odds: A 60–40 Race
The Senate is where the uncertainty lives. Republicans hold 53 seats. Because Vice President J.D. Vance breaks ties for Republicans, Democrats need 51 seats for control, which is a net gain of four.
Four is a lot in a single cycle, but the map is unusually friendly to Democrats for a class that was supposed to favour Republicans. Several Republican incumbents retired, including Thom Tillis in North Carolina and Joni Ernst in Iowa. Texas nominated Ken Paxton, who took the Republican nomination from incumbent John Cornyn. Ohio holds a special election for the seat Vance vacated, with former senator Sherrod Brown on the ballot.
The market prices all of that at roughly 60% for a Democratic Senate. Polymarket shows 60.5¢; Kalshi quotes 59¢ bid and 60¢ ask. Polymarket's Senate market has traded about $4.8 million, and Kalshi about 9.3 million contracts across both sides. That is thinner than the House market, which matters if you plan to trade size: a few hundred thousand dollars moves this book more than it moves the House. Our guide to prediction market liquidity explains how to judge depth before you trade.
The Senate Races That Decide Control
Chamber control is just the sum of individual races, and Kalshi lists a market for nearly every one of them. These are the seats that matter, with Kalshi's bid and ask on the leading candidate as of September 21.
| State | Seat now held by | Race | Kalshi leader (bid / ask) |
|---|---|---|---|
| North Carolina | Republican (open) | Roy Cooper (D) vs Michael Whatley (R) | Cooper 92¢ / 93¢ |
| Georgia | Democrat | Jon Ossoff (D) vs Mike Collins (R) | Ossoff 93¢ / 94¢ |
| Minnesota | Democrat (open) | Peggy Flanagan (D) vs Michele Tafoya (R) | Flanagan 90¢ |
| New Hampshire | Democrat (open) | Chris Pappas (D) vs John E. Sununu (R) | Pappas 82¢ / 83¢ |
| Alaska | Republican | Democratic side vs Republican side | Democrats 70¢ / 71¢ |
| Nebraska | Republican | Dan Osborn (I) vs Pete Ricketts (R) | Ricketts 68¢ / 69¢ |
| Maine | Republican | Troy Jackson (D) vs Susan Collins (R) | Jackson 67¢ / 68¢ |
| Michigan | Democrat (open) | Abdul El-Sayed (D) vs Mike Rogers (R) | El-Sayed 66¢ / 67¢ |
| Iowa | Republican (open) | Josh Turek (D) vs Ashley Hinson (R) | Hinson 60¢ / 61¢ |
| Texas | Republican | James Talarico (D) vs Ken Paxton (R) | Talarico 59¢ / 60¢ |
| Ohio (special) | Republican | Sherrod Brown (D) vs Jon Husted (R) | Brown 56¢ / 57¢ |
| Kansas | Republican | Adam Hamilton (D) vs Roger Marshall (R) | Marshall 73¢ / 74¢ |
| Florida (special) | Republican | Angie Nixon (D) vs Ashley Moody (R) | Moody 88¢ / 89¢ |
Our 2026 midterms trading guide explains how these single-race markets are structured and where each venue lists them. Live pages for eight of these races — including Texas, Maine and the Ohio special election — track both venues side by side. Texas is the most traded race on the board by a distance, with about 13.5 million contracts across both candidates. It is also one of the most consequential: a Democratic win in Texas would be the first in a Senate race there since 1988, and the market gives it better than even odds.
If every favourite wins, Democrats take five Republican-held seats — North Carolina, Alaska, Maine, Texas and Ohio — while holding everything they currently have. That would produce a 52–48 Senate. The problem for anyone buying Democratic control is that three of those five flips are priced between 56¢ and 70¢, which is a long way from certain.
How the Race Prices Add Up to 60%
They add up, and checking is a useful exercise. Treat each race as an independent coin weighted by its Kalshi price and add up the expected seat changes. The competitive races in the table, plus the safe Democratic seats in Virginia, Colorado, Illinois and New Mexico, produce an expected Democratic gain of about 3.6 seats, and a 52% chance of reaching the net gain of four.
That is lower than the 60% the chamber market shows, so the next step is to add the seats nobody talks about. Long shots in states like South Carolina (Democrats around 13¢), Mississippi (around 8¢) and Arkansas (around 6¢) each contribute a fraction of a seat. Add them all and the expected gain rises to about 3.8 seats, with a 59% chance of reaching four — within a point of the chamber market. Count Dan Osborn's Nebraska seat with the Democrats and it rises to 65%, which is why his race is the swing assumption in any Senate model.
Two lessons come out of this. First, the race markets and the chamber market are consistent, which means there is no obvious mispricing to trade between them. Second, the independence assumption understates risk. Races move together when the national environment shifts, so the real distribution is wider than the calculation suggests, in both directions. A good night for either party does not flip one race; it flips four.
This kind of cross-checking is the everyday work of pricing, and it is exactly what our expected value guide and EV calculator are built for.
Balance of Power: The Combined Market
Polymarket also lists a single market on the combined result, which has traded about $12.9 million. The Democratic sweep leads at 60.5¢, a Republican Senate with a Democratic House follows at 30.5¢, and a Republican sweep trails at 7.5¢. The fourth combination — a Democratic Senate with a Republican House — carries only a sliver, because it would require Democrats to win the harder chamber while losing the easier one.
Notice that the Democratic sweep price equals the Democratic Senate price. That is the market saying that almost every path to a Democratic Senate also runs through a Democratic House, which is consistent with the 91% House price. Anyone holding a view on the Senate is, in practice, trading the same exposure in either market.
The combined market is useful for one specific trade: a bet on divided government. At 30.5¢, a Republican Senate and Democratic House is the natural hedge for anyone whose business or portfolio depends on gridlock rather than a unified Democratic agenda. Our guide to hedging with prediction markets works through that kind of position with numbers.
Why Polymarket and Kalshi Agree So Closely
On the House, the two venues sit within half a cent. On the Senate, within one cent. On individual races, they generally match as well. That is not true of every market we track — the 2028 party market currently shows a six-point gap between the two venues, which our arbitrage scanner flags as a live opportunity.
The midterm markets are close for a simple reason: they are liquid, close to resolution, and full of professional traders who arbitrage any gap down to the cost of doing it. Kalshi's fee is 0.07 × contracts × price × (1 − price), which at 60¢ costs about 1.7¢ per contract. Polymarket's taker fee runs 0.75–1.80% depending on market category. Moving money between a dollar account and a USDC wallet costs something too. A gap smaller than all of that is not a gap anyone can profit from, so it does not survive.
If you want to see how that math works for any pair of prices, our market fee calculator runs both venues' formulas, and our arbitrage guide covers the full workflow.
How Midterm Markets Resolve
Election markets settle on certified results, not on election night calls, and that gap can run for weeks. Read each rulebook before you buy, because the details decide close contracts.
Three quirks matter in 2026. Maine and Alaska both use ranked-choice voting in general elections, so a race without a first-round majority goes to additional tabulation rounds that can take days or more. Georgia requires a majority to win outright and sends a race without one to a runoff. And some states count mail ballots slowly — Alaska and California are routinely among the last to finish — which leaves House control uncertain for days if the chamber is close.
Each venue defines what counts as a win and which sources it relies on, and the two do not always use identical wording. That is the single biggest risk in cross-venue trading on elections: two contracts that look identical can settle differently on an edge case. Our explainer on how prediction markets resolve covers what to look for in the rules.
What the Odds Do Not Tell You
Markets are well calibrated on average, and prediction markets have a respectable record on elections. They also have well-known misses. In 2016 they had Hillary Clinton as a strong favourite; in 2022 they priced a Republican wave that arrived as a ripple, with Democrats holding the Senate. A 91% favourite loses about one time in eleven, and a 60% favourite loses four times in ten. Neither outcome is a failure of the market, but both are expensive if you sized as though they were certain.
The second limit is that prices reflect the people trading, not the electorate. Election markets draw a particular crowd, and there is evidence of partisan money leaning into some contracts. That is part of why we compare two venues: a price that holds on both, with different users on each, is harder to dismiss as one crowd's enthusiasm. Our guide to prediction market accuracy covers the research.
How to Trade 2026 Midterm Odds
Most US traders will use Kalshi, which is regulated by the CFTC as a designated contract market, settles in dollars and pays about 4% on idle cash while you wait for November. Polymarket offers the same markets with deeper history and a larger international user base, settling in USDC on Polygon. Robinhood offers election contracts inside its brokerage app through exchange partners — see our Robinhood review. If you are choosing a venue, our ranking of the best prediction market apps compares them on fees, access and depth, and Polymarket vs Kalshi goes head to head.
The midterms trading guide linked above covers the step-by-step, including legal access and position sizing. Two points bear repeating. First, election contracts have been at the centre of the fight over whether these markets are gambling, and state positions differ — check our legal guide for where you live. Second, profits are taxable whether or not you withdraw them, and Kalshi issues a 1099-MISC; see our prediction market tax guide.
Key Dates Before the Midterms
| Date | Event | Why it matters for prices |
|---|---|---|
| October 2026 | Early and mail voting underway in most states | Early-vote data moves race markets |
| November 3, 2026 | Election Day | Most volatility of the cycle |
| Days to weeks after | Counting, recounts, ranked-choice tabulation | Close races stay open |
| December 2026 | State certification deadlines | Markets settle on certified results |
| January 3, 2027 | New Congress sworn in | Control becomes official |
The single biggest mistake in election trading is holding a large position through a volatile night without a plan for what you will do if the early count breaks against you. Decide your exit before the polls close, not after, and run the numbers on a winning scenario through our take-home calculator so fees and tax do not surprise you.
Frequently Asked Questions
Who is favored to win the House in 2026?
Democrats. As of September 21, 2026, they trade at 91.5¢ on Polymarket and 91.3–91.5¢ on Kalshi, which implies about a 91% chance of winning the House. The president's party has lost House seats in all but three midterms since the Civil War.
What are the odds Democrats win the Senate in 2026?
About 60%. Polymarket shows 60.5¢ and Kalshi quotes 59¢ bid, 60¢ ask. Democrats need a net gain of four seats because Vice President J.D. Vance breaks ties for Republicans.
Which Senate races will decide control?
Maine, North Carolina, Texas, Ohio and Alaska are the Republican-held seats where Democrats are favoured, while Michigan and New Hampshire are the most vulnerable Democratic seats. Texas is the most heavily traded race on Kalshi.
Are prediction markets accurate for midterm elections?
They are well calibrated on average but not infallible — they overestimated Republicans in 2022. Treat a 60% price as a 60% chance, not a forecast of the result. Our accuracy guide covers the evidence.
Can I legally bet on the midterms in the US?
Kalshi offers election contracts to US users as a CFTC-regulated exchange, and Robinhood offers them inside its app through exchange partners. Some states contest these markets, so check our legal guide for your state.
When do midterm prediction markets pay out?
On certified results, not election night projections. Close races, ranked-choice tabulation in Maine and Alaska, and runoffs can delay settlement for days or weeks.
Why do Polymarket and Kalshi show the same midterm odds?
Because the markets are liquid and professional traders close any gap larger than the fees. Kalshi's fee at 60¢ is about 1.7¢ per contract, so a one-cent difference is not worth trading.
Do I pay tax on midterm betting winnings?
Yes. Profits are taxable in the year they are realised, and Kalshi issues a 1099-MISC. See our prediction market tax guide for the three treatments in use.
