Our final calls for the 2026 midterms
Twenty-eight published calls, one baseline formula, and two clearly labeled editorial overrides. Here is where we think the Senate, the House and the closest governorships land on November 3.
These are predictions, not preferences. Every race runs through the same formula: independent ratings, polling averages and prediction-market prices, each put on the same scale and averaged. Two calls are ours, not the formula’s. In Kansas and Iowa the formula leaned Republican by a hair, less than the polling averages disagree with each other, and our editors called both for the Democrat. We label them on the scoreboard and the race cards and show what the formula said.
The confidence column matters as much as the pick. A Toss-up call is a coin flip we were willing to make; a Strong call means all three signals point the same way by a wide margin.
Our boldest callsKansas and Iowa for the Democrats. Our formula wouldn’t make either call.Both lean Republican on the formula by less than the polling averages disagree. Here’s the case for, the case against, and why control doesn’t hinge on either. Read why →The scoreboard
| Office | Race | Our call | Confidence |
|---|---|---|---|
| Senate | Iowa | Josh Turek (D)pickup Editor’s call | Toss-up |
| Senate | Kansas | Adam Hamilton (D)pickup Editor’s call | Toss-up |
| Senate | Maine | Troy Jackson (D)pickup | Lean |
| Senate | Alaska | Mary Peltola (D)pickup | Lean |
| Senate | Texas | James Talarico (D)pickup | Lean |
| Senate | Ohio (special) | Sherrod Brown (D)pickup | Likely |
| Senate | Michigan | Abdul El-Sayed (D) | Likely |
| Senate | Nebraska | Pete Ricketts (R) | Likely |
| Senate | South Carolina | Darline Graham (R) | Likely |
| Senate | Minnesota | Peggy Flanagan (D) | Strong |
| Senate | New Hampshire | Chris Pappas (D) | Strong |
| Senate | Georgia | Jon Ossoff (D) | Strong |
| Senate | North Carolina | Roy Cooper (D)pickup | Strong |
| Governor | Georgia | Keisha Lance Bottoms (D)pickup | Toss-up |
| Governor | Nevada | Aaron Ford (D)pickup | Lean |
| Governor | Texas | Greg Abbott (R) | Likely |
| Governor | Wisconsin | David Crowley (D) | Likely |
| Governor | Kansas | Ty Masterson (R)pickup | Likely |
| Governor | Ohio | Amy Acton (D)pickup | Likely |
| Governor | Alaska | Jonathan Kreiss-Tomkins (D)pickup | Strong |
| House | TX-35 | Johnny Garcia (D) | Lean |
| House | NY-17 | Cait Conley (D)pickup | Lean |
| House | OH-09 | Marcy Kaptur (D) | Likely |
| House | FL-14 | Kathy Castor (D) | Likely |
| House | NC-01 | Don Davis (D) | Likely |
| House | MI-07 | William Lawrence (D)pickup | Likely |
| House | PA-08 | Paige Cognetti (D)pickup | Likely |
| House | TX-15 | Bobby Pulido (D)pickup | Strong |
More of our calls
Our race-by-race series also called these contests. Each has its own page with the polls, the money and what would make us wrong.
| Office | Race | Our call |
|---|---|---|
| Governor | Iowa | Rob Sand (D)pickup |
| House | AZ-01 | Amish Shah (D)pickup |
| House | PA-07 | Bob Brooks (D)pickup |
| House | PA-10 | Janelle Stelson (D)pickup |
| House | VA-01 | Shannon Taylor (D)pickup |
| House | WA-03 | Marie Gluesenkamp Perez (D) |
| House | PA-01 | Brian Fitzpatrick (R) |
Every race goes through the same formula, so no race gets special treatment. We put three independent signals on the same scale, from −3 (strongly Democratic) to +3 (strongly Republican), and average them:
- Race ratings from Cook Political Report, Sabato's Crystal Ball and Inside Elections: Safe 3, Likely 2, Lean 1, Tilt 0.5, Toss Up 0.
- Polling average: the leader's margin divided by 2, capped at 3. Where no average exists we use nonpartisan public polls only; party and campaign polls are excluded.
- Prediction markets: the average of Kalshi and Polymarket, with the gap from 50% divided by 10, capped at 3.
The sign of the average is the call. Its size sets the confidence: 2 or more is Strong, 1 or more Likely, 0.4 or more Lean, anything closer Toss-up. When our editors override the formula, the race card says so and shows what the formula said.
What could make us wrong
- Polling misses. Ohio polls understated Republicans by about 6 points in 2018, according to Sabato's Crystal Ball, and several of our Democratic calls rest on leads smaller than that.
- Markets. In the House toss-ups especially, prediction markets lean further toward Democrats than ratings or polls do. If the markets are wrong, several of those calls move.
- Stale averages. Some averages end in mid-October and miss late polls. We note it on each race card where it matters.
- Late rating changes. Forecasters are still moving races this week. We will update these calls before Election Day if the numbers change.
On election night, follow the live dashboard for race calls as they come in.