2026 Midterm Prediction Markets: Where Kalshi and Polymarket disagree

Kalshi and Polymarket are pricing many of the same 2026 US midterm prediction markets, but they don't always agree on identical figures.

On October 2, 2026, the difference between Kalshi and Polymarket.com was generally small. Across 219 Senate and governor outcomes priced on both platforms in iPredicta's own matched data, the typical (median) gap between the two venues' mid prices was under one percentage point, and 79% of matched outcomes were within two points of each other. The average gap was around 1.3 points for governor races and around 1.7 points for Senate races.

The more pertinent question then is not so much why Kalshi and Polymarket disagree, as most of the time, they don't. It is when they do disagree, how large those differences are, and what is actually causing them.

The 2026 midterm prediction markets provide some useful examples. In the Georgia governor's race on October 2, Polymarket.com priced a Democratic win at 47.5% against 46.5% on Kalshi, and a Republican win at 53.5% against 51.5%, in a race sitting close to the 50-50 mark. A day earlier, the two had briefly disagreed on who was favourite.

New Hampshire presents a different puzzle. Polymarket.com was pricing a Republican victory at around 95%, while Kalshi's buyers and sellers were quoting between 90% and 95%. On the surface, that looks like an even larger disagreement. But both order books were thin, and very little had traded on either in the previous day.

The distinction matters and helps explain How Prediction Markets Work and Why Some Prediction Market Contracts Are Thin.

A difference between two prediction market prices can reflect genuinely different assessments of an election. But it can also be produced or exaggerated by liquidity, bid-ask spreads, fees, contract wording or settlement rules.

This 2026 midterm prediction market analysis examines the largest and most interesting differences, using Georgia and New Hampshire as contrasting case studies. The aim is not simply to identify which platform has the higher number, but to understand why the numbers diverge, and what that tells us about how prediction markets should be read.

For the latest 2026 Senate, House and Governor prediction market prices, see iPredicta's US midterms prediction markets page.

What 2026 midterm prediction markets do Kalshi and Polymarket agree on?

The broad picture across the 2026 midterms is one of convergence rather than disagreement.

The clearest example is the Senate. As of October 2, Polymarket.com priced Democratic control at 62.5%, with Kalshi at 61.5%. Both markets therefore put the probability of a Democratic Senate at roughly three in five.

The House shows a similarly close relationship. Polymarket.com priced Democratic control at 92.5%, compared with 91.5% on Kalshi, a difference of just one percentage point.

This close relationship is not limited to the headline control markets. Looking across individual races, most matched outcomes on the two platforms sit within a point or two of each other.

That agreement is telling. These are separate exchanges, with different traders, order books and market mechanics, yet when the underlying information points in broadly the same direction, their prices tend to converge.

It also means that a difference between the two platforms should not automatically be interpreted as one market having identified something the other has missed.

The more useful question is whether a price gap is persistent, and what it actually reflects.

Georgia governor prediction markets: A genuine difference between Kalshi and Polymarket?

Georgia is one of the best examples of how a difference between the two prediction markets can open and close.

As of October 2, Polymarket.com priced Democrat Keisha Lance Bottoms at 47.5%, against 46.5% on Kalshi, and Republican Rick Jackson at 53.5%, against 51.5%. The wider difference sat on Jackson's side of the race. With both markets close to 50-50, a gap of that size carries more weight than the same gap would in a race priced above 90%.

The closeness of the election is consistent with the picture coming from polling. A Big Data Poll conducted September 21-23 found Jackson leading Bottoms 48.7% to 46.7% among likely voters, narrowing to 50.5% to 49.5% once undecided voters' leanings were allocated. An InsiderAdvantage poll of 1,200 likely voters conducted September 22-23 put Jackson ahead 48% to 46%, with 6% undecided.

Neither Georgia governor prediction market was presenting a radically different view of the race. Both were pricing a close contest with Jackson narrowly favoured.

How persistent is the Georgia prediction market gap?

The gap between the trading platforms has changed considerably over time. On September 11, Polymarket.com and Kalshi both priced a Democratic victory at around 56.5%, and the following day's prices were within a point of each other.

On October 1, the picture was briefly different. Polymarket.com's price for Bottoms moved as high as 52.5% during the day, making her the favourite there, while Kalshi's price for her never traded above 50%. By the end of the day Polymarket.com had fallen back to 48.5%, in line with Kalshi, and on October 2 the two were within a point of each other on her price.

Both markets have been moving, with the distance between them changing over time. Kalshi's Democratic price traded as low as 43% on June 28 and as high as 66% on April 27, and in September ranged between 49% and 61%, ending the month near the bottom of that range.

The Georgia market is not simply a case of one platform being consistently more bullish on one candidate than the other. The gap between the two platforms has changed in both size and direction as the market has moved.

What does the Georgia price gap mean?

The size of the Georgia gap looks different when viewed alongside the trading activity behind the two markets.

On October 2, Polymarket.com's Georgia market had generated around $140,000 in lifetime trading volume, about three times the lifetime volume of its New Hampshire market. Kalshi's two Georgia contracts had traded around 1.15 million contracts between them, nearly eight times as many as its New Hampshire contracts. The two platforms measure volume in different units, so the Georgia figures cannot be compared with each other.

A difference in headline prices is a difference between two quotes on two separate exchanges, each with its own traders, order book, fees and contract rules. It shows that the two sets of traders are, at that moment, settling on slightly different numbers. It does not, by itself, show that either set of traders has reached a different conclusion about the election.

It means a price gap can be real without telling us much about the race itself.

In Georgia, the evidence points to a genuine difference in the prices being quoted by the two markets, but not to a fundamentally different assessment of the election. Both were clustered around the same close-race view.

Why Kalshi and Polymarket disagree on Georgia

The most interesting feature of the Georgia market is not the size of the gap but how quickly it moves. On October 1, for part of the day, one platform had Bottoms as favourite while the other had Jackson. A day later they were within a point of each other on Bottoms and two points apart on Jackson.

That is happening in a race where the available polling offered little basis for a decisive market view. Recent polls had the candidates within a few points of each other. Big Data Poll's two-point gap among likely voters was within its stated sampling error of plus or minus 4.0 points; InsiderAdvantage did not publish a margin of error for its Georgia figures. The underlying race was therefore genuinely uncertain, leaving relatively little separation for the two markets to price.

There is no clear evidence that one platform reacted to a specific piece of information before the other. That makes it difficult to argue that the gap represented a fundamentally different assessment of the election.

A more straightforward explanation is the nature of the markets themselves. Kalshi and Polymarket are separate exchanges, so the traders, available liquidity and outstanding orders are not identical. In a close race, relatively modest differences in buying and selling pressure can move the quoted price on one platform without immediately producing the same move on the other. Liquidity can amplify those movements, particularly where a market has less depth, but there is no evidence that liquidity alone caused the Georgia gap. Both Georgia markets had far more trading activity behind them than their New Hampshire counterparts.

The historical data supports that interpretation. The gap between the platforms has changed repeatedly rather than showing a consistent advantage for one side.

That does not mean the difference is meaningless. It means the difference needs to be interpreted as a gap between two market prices, rather than two completely independent forecasts of the election.

Georgia therefore provides a useful warning about reading prediction market probabilities too literally. A gap of a few points between two highly uncertain markets can look substantial, but it does not necessarily mean the traders on one platform have discovered something the traders on the other have missed.

Learn more about the differences between Kalshi and Polymarket prediction markets in our Kalshi vs Polymarket guide.

New Hampshire governor prediction markets: A bigger gap, but weaker signal

New Hampshire provides a stark contrast with Georgia. The gap between the headline prices looks larger, but both markets are considerably thinner.

Polymarket.com priced a Republican victory in the New Hampshire governor's race at around 95%. On Kalshi, buyers were bidding 90% and sellers asking 95%, a wide range that is itself a sign of a thin market. On the surface, that looks like a more significant disagreement than the gap in Georgia.

But the evidence behind both prices is limited. Polymarket.com's New Hampshire market had generated around $41,000 in lifetime volume, with only around $100 traded in the previous 24 hours. Kalshi's two New Hampshire contracts had traded around 150,000 contracts between them, around 31,000 of them on the Republican side, and none on that side in the previous 24 hours. Polymarket.com's best bid, at 95%, was also level with Kalshi's best offer, so the two books were touching rather than far apart.

That makes it much harder to interpret the difference as evidence that traders on the two platforms have fundamentally different views of the race.

Instead, the New Hampshire example demonstrates why the headline probability needs to be considered alongside the depth of the market producing it. A price can move several percentage points without representing a major shift in the underlying consensus if there is relatively little trading activity supporting the quoted price.

This does not make either price meaningless. It simply means that a 95% probability in a thin market should not necessarily be treated in the same way as a similarly priced outcome backed by a much deeper order book.

In Georgia, the two platforms were pricing a genuinely uncertain race slightly differently, with far more trading activity behind both prices than in New Hampshire. In New Hampshire, the apparent disagreement looks larger, but the evidence behind both prices is far thinner.

What do differences between Kalshi and Polymarket tell us?

The 2026 midterm prediction markets show that disagreement between Kalshi and Polymarket is usually smaller than the headline numbers might imply.

Across the wider market, the two platforms are generally closely aligned. Where they do diverge, the examples from Georgia and New Hampshire show why the size of the gap alone is not enough to explain what is happening.

Georgia showed two markets, each far busier than its New Hampshire counterpart, pricing an extremely close race slightly differently, at one point even disagreeing on the favourite, but with little evidence that either had identified a truly different view of the election.

New Hampshire presented the opposite situation. The headline gap looked larger, but both prices came from thin markets, so the difference was less compelling as evidence of genuine disagreement, with both platforms already heavily favouring a Republican victory.

For anyone using prediction markets to assess the 2026 midterms, the most interesting signal may therefore not be where Kalshi and Polymarket disagree, but why they disagree when they do.

For the latest 2026 Senate, House and Governor prediction market prices, see iPredicta's US midterms prediction markets page, or read our broader 2026 midterm election prediction markets analysis for an overview of the markets for control of Congress.

Market snapshot: 09:15 UTC, 2 October 2026. Prices are the midpoint between the best bid and best offer on each platform. Prices and trading volumes can change as trading continues.

READ MORE

• What lessons can we take from the 2024 election for the 2026 midterms?
• What developments could move the 2026 midterm prediction markets?
• 5 things traders are telling us about the 2026 midterm prediction markets

Polymarket figures in this article refer to markets on Polymarket.com unless otherwise stated. Polymarket.com and Polymarket US operate separate order books, so prices can differ between the two platforms.