Election night
Live results and projections for November 3, 2026. Returns are unofficial until certified.
Pre-election forecast of 12:24 pm CT · checks every 30 seconds
Feed details
Displayed model run: 12:24 pm Central, Nov 3. Received feed’s official source last checked: not recorded. This browser is checking for the latest feed. Checks every 30 seconds while this tab is visible and online.
Showing model run 12:24 pm Central, Nov 3.No returns yet. Until counties report, these are the pre-election forecasts.
Polls close at 7 pm Central. Once the first counties report, every figure on this page moves with the count. The Secretary of State’s results have the raw count.
Results and projections
Amendment 3
(HCS HJR 73): repeal 2024 reproductive-rights amendment; abortion only for emergencies, fetal anomalies, rape/incest under 12 weeks; parental consent; ban gender-transition procedures for minorsNo votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: Passes 30%, Fails 70%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
Amendment 3: why the live figure is what it is Part of the expected yes share Points of yes share Polls’ yes share (43.0% × 47%) 20.3 pts Undecided voters (13.7% × 0.22 × 47%) 1.4 pts Starting estimate (45.7% × 53%) 24.2 pts Expected yes share 45.9% - Polls average 43.0% yes, 43.3% no and 13.7% undecided (3 polls; the latest, Remington Research Group/Missouri Scout Oct 15, carries 73% of the weight).
- In past Missouri measures, results ended up about 0.22 times the undecided share above the final-poll yes. That is an average correction from past measures, not a measured split of this measure’s undecided voters. It puts the polls at 46.1%.
- Our starting estimate is 45.7%: the 2024 abortion-rights amendment’s vote, reweighted to the expected 2026 electorate (51.7% keep rights) and moved by the same statewide swing as the seat model (0.784 × 3.3 points), gives repeal 45.7%, with no step toward the polls, which count separately (a judgment, decision D-M10).
- The model weighs the polls at 47% and the starting estimate at 53% (the polls’ combined weight, 35.8, against 40 for the starting estimate), so the expected yes share is 45.9%.
Eight times in ten the simulated yes share lands between 34.8% and 57.0%. The range allows for the polls and the starting estimate both being off, and for polls missing in the same direction on every measure.
Yes clears 50% in about 30,390 of 100,000 simulated elections.
What the model does not know. Campaign money and betting markets are not inputs; the starting estimates are our judgment, written down in the decisions log.
Amendment 6
(initiative petition): initiative/referendum a fundamental right; legislature needs 80% of both chambers to refer changes to voter-approved statutesNo votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: Passes 70%, Fails 30%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
Amendment 6: why the live figure is what it is Part of the expected yes share Points of yes share Polls’ yes share (46.5% × 47%) 21.9 pts Undecided voters (28.8% × 0.22 × 47%) 3.0 pts Starting estimate (56.0% × 53%) 29.6 pts Expected yes share 54.5% - Polls average 46.5% yes, 24.7% no and 28.8% undecided (2 polls; the latest, Remington Research Group/Missouri Scout Oct 15, carries 73% of the weight).
- In past Missouri measures, results ended up about 0.22 times the undecided share above the final-poll yes. That is an average correction from past measures, not a measured split of this measure’s undecided voters. It puts the polls at 52.9%.
- Our starting estimate is 56.0%: voters rejected the Republican initiative curbs 80% to 20% in August 2026, and polls show Yes well ahead with about 30% undecided, but this is a complex structural amendment opposed by the governor and Republican leaders, and undecided voters have historically broken toward No.
- The model weighs the polls at 47% and the starting estimate at 53% (the polls’ combined weight, 35.6, against 40 for the starting estimate), so the expected yes share is 54.5%.
Eight times in ten the simulated yes share lands between 42.9% and 66.2%. The range allows for the polls and the starting estimate both being off, and for polls missing in the same direction on every measure.
Yes clears 50% in about 70,160 of 100,000 simulated elections.
What the model does not know. Campaign money and betting markets are not inputs; the starting estimates are our judgment, written down in the decisions log.
Amendment 7
(SS SCS SJR 95): Show-Me Prosperity Fund, permanent endowment to replace state taxesNo votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: Passes 27%, Fails 73%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
- No public poll. The expected yes share, 42.0%, is our starting estimate alone: there is no poll and no campaign on either side; the ballot title’s appeal (“instead of taxing Missouri residents”) is set against voters rejecting the income-tax phase-out 83% to 17% in August 2026 and a 2022 treasurer-investment amendment that took 45.7%; low-information legislative referrals in Missouri have split.
- The range is wide because unpolled Missouri referrals have varied widely (D-M6).
Eight times in ten the simulated yes share lands between 24.0% and 60.0%.
Yes clears 50% in about 27,260 of 100,000 simulated elections.
What the model does not know. Campaign money and betting markets are not inputs; the starting estimates are our judgment, written down in the decisions log. There is no public poll.
Amendment 8
(CCS HCS SS SJR 87): constitutionally require elected county sheriffs; removal only by quo warranto from the Attorney GeneralNo votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: Passes 73%, Fails 27%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
- No public poll. The expected yes share, 58.0%, is our starting estimate alone: there is no poll and no organized opposition; the ballot title writes existing practice into the constitution; law-enforcement referrals have ranged from 39% to 63% yes, and the August 2026 elected-assessor referral passed with 76.9%.
- The range is wide because unpolled Missouri referrals have varied widely (D-M6).
Eight times in ten the simulated yes share lands between 40.0% and 76.1%.
Yes clears 50% in about 73,060 of 100,000 simulated elections.
What the model does not know. Campaign money and betting markets are not inputs; the starting estimates are our judgment, written down in the decisions log. There is no public poll.
Proposition A
veto referendum on HB 1 (2025 Second Extraordinary Session) congressional mapNo votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: Passes 20%, Fails 80%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
Proposition A: why the live figure is what it is Part of the expected yes share Points of yes share Polls’ yes share (35.1% × 43%) 15.1 pts Undecided voters (32.9% × 0.22 × 43%) 3.1 pts Starting estimate (43.0% × 57%) 24.5 pts Expected yes share 42.7% - Polls average 35.1% yes, 32.1% no and 32.9% undecided (2 polls; the latest, SLU/YouGov Aug 24, carries 99% of the weight).
- In past Missouri measures, results ended up about 0.22 times the undecided share above the final-poll yes. That is an average correction from past measures, not a measured split of this measure’s undecided voters. It puts the polls at 42.3%.
- Our starting estimate is 43.0%: Yes approves the 2025 map and No keeps the 2022 map; the February SLU poll had the map 41% to 44%, and the 2018 veto referendum lost with 32.5%; a Republican-leaning state and a favourable ballot title pull the other way.
- The model weighs the polls at 43% and the starting estimate at 57% (the polls’ combined weight, 30.3, against 40 for the starting estimate), so the expected yes share is 42.7%.
Eight times in ten the simulated yes share lands between 31.0% and 54.5%. The range allows for the polls and the starting estimate both being off, and for polls missing in the same direction on every measure.
Yes clears 50% in about 20,210 of 100,000 simulated elections.
What the model does not know. Campaign money and betting markets are not inputs; the starting estimates are our judgment, written down in the decisions log.
State Auditor
The only statewide partisan race on the ballotNo votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: expected Democratic share 43.1%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
State Auditor: why the live figure is what it is Part of the expected share Effect 2024 Democratic share in the down-ballot statewide races 39.9% Statewide swing in the legislative forecast +3.3 pts MO-2 poll update to the statewide mood +0.3 pts Senate-district poll update to the statewide mood +0.1 pts Republican incumbent’s money and name −1.0 pts Auditor polls, candidate effect +0.5 pts Expected Democratic share 43.1% - Last cycle’s down-ballot statewide races (Secretary of State, Treasurer, Attorney General and Lieutenant Governor) gave Democrats 39.9%, weighted to the expected 2026 turnout. Add the same statewide swing as the legislative forecast (+3.3 pts, plus +0.3 pts from the MO-2 polls and +0.1 pts from two Senate-district polls) and take off 1.0 pts for the Republican incumbent’s money and name and add 0.5 pts from two Auditor polls (SLU/YouGov in August, Remington for Missouri Scout in October), read as evidence on the candidates, not the statewide mood: 43.1%.
- The range combines 1.5 points of uncertainty in the statewide swing (decision D-03) with a 1.4-point candidate effect. Only the 1.5-point candidate effect before the polls, the 1.0-point incumbent adjustment and the polls’ 4-point error are our judgment (D-24, D-26).
Eight times in ten the simulated Democratic share lands between 40.0% and 46.0%.
What the model does not know. Two Auditor polls inform the candidate effect (D-26); the 1.0-point incumbent adjustment and the polls’ 4-point error are our judgment.
U.S. Representative, District 2
US House, 2nd District: Fred Wellman (D) v Ann Wagner (R)No votes counted yet
Awaiting returns. The chance and range are the pre-election forecast; a projection from the count appears once 5% of the expected vote is counted.
Before returns: Democrat 44%, Republican 56%
Why?
Under 5% of the expected vote is counted, so this shows the pre-election forecast unchanged.
U.S. Representative, District 2: why the live figure is what it is Part of the expected share Effect Fundamentals (48.4% × 33%) 15.9 pts Polls (50.5% × 51%) 25.7 pts Betting markets (49.0% × 16%) 8.0 pts Centre of the blend 49.6% - Fred Wellman (D) is expected to take 49.6% of the two-party vote against Ann Wagner (R). The model blends its estimates, each with its weight.
- Before any poll, the model expects Wellman to take 48.4%. The national generic ballot is 54.2% of the two-party vote (Democrats lead the national generic ballot by 8.4 points), and in 2024 this district voted 3.3 pts less Democratic than the nation for president. Incumbency is worth 2.0 pts to the Republican and the spending gap 0.5 to the Republican.
- 4 polls, adjusted for partisan sponsors, average 50.5%. Together they count as 1.6 fresh 400-person polls.
- Betting markets price Wellman at 40.2% to win, which the model reads as a 49.0% share, counted at half a poll’s weight. Only freshly priced markets count (D-C7).
- The blend is 33% fundamentals, 51% polls and 16% betting markets, which centres the expected Wellman share at 49.6%.
- The 4 outside ratings shown on this page are for comparison; the model does not use them.
Eight times in ten the simulated Wellman share lands between 45.5% and 53.7%.
Wellman finishes ahead in about 44,200 of 100,000 simulated elections.
What the model does not know. Ratings are not inputs; markets count only when freshly priced (D-C7).
Each strip is the contest’s projected final share of the two-way vote: yes over yes plus no for a measure, Democrat over Democrat plus Republican for the auditor. The band is the range the middle 80% of simulations fall in (10th to 90th percentile), the tick is the projected share, and the heavy rule is 50%. The chance mark splits the simulations: the share in which the yes side, or the Democrat, finishes above 50%, and the share in which it does not. Votes counted are the unofficial count; the Missouri Secretary of State has the official results.
Calls
No race has been projected yet. A race is called when the live model gives one outcome at least a 99% chance with at least 15% of its expected vote counted; a call is withdrawn if that chance falls below 97.5%.
The two chambers
- pre-election forecastRepublican supermajority in the House: chance of each outcomeDemocrats expected to win 60 of 163 seats, 80% range 54 to 66. Republicans need 109 of 163 seats for two-thirds; Democrats break it by holding them under 109.
- pre-election forecastRepublican supermajority in the Senate: chance of each outcomeDemocrats expected to win 11.5 of 34 seats, 80% range 10 to 12. Republicans need 23 of 34 seats for two-thirds; Democrats break it by holding them under 23.
Pre-election forecast of 12:24 pm Central, Nov 3; the live model takes over once counties report. Expected seats count every member of the chamber, including the 17 senators not up this year; the range is the 10th to 90th percentile.
Follow a district on election night
Choose a House or Senate district to see its live chance and whether it has returns yet. District pages hold the pre-election forecast.
Manage or share saved races · Chamber outcomes · Results and projections
Where the count stands
| Region | Counted | Counted yes share | Expected | Projected |
|---|---|---|---|---|
| St. Louis area | None yet | — | 36.5% | 36.5% |
| Kansas City area | None yet | — | 34.4% | 34.5% |
| Other cities | None yet | — | 50.5% | 50.5% |
| Rural | None yet | — | 63.1% | 63.1% |
Each counted figure names its basis: precincts is precincts reporting, est. vote is votes counted over votes expected, votes is the absentee batch plus the precincts reporting since (an estimate of the vote counted), reported is the share the election authority gave, final count is a hand-flagged complete count, and unknown means the source did not say. Counted is votes counted over the region’s expected votes so far. Expected is the yes share the region was expected to give before any returns. Regions: St. Louis area is St. Louis City and St. Louis, St. Charles, Jefferson, Franklin, Lincoln and Warren counties; Kansas City area is Kansas City and Jackson, Clay, Platte, Cass and Ray; other cities are Greene, Christian, Boone, Cole, Callaway, Jasper, Newton, Buchanan and Cape Girardeau counties (Springfield, Columbia, Jefferson City, Joplin, St. Joseph and Cape Girardeau); rural is the rest.
County by county
No county has reported yet for Amendment 3.
Legislative seats
Projected winners
No projected winners yet. A seat is listed here once the live model gives a candidate at least a 97.5% chance, and is marked called at 99% with at least 15% of its vote counted. Either can come off again if later counts shift.
House seats in doubt
| District | Candidates | Chance before returns | Chance now |
|---|---|---|---|
| House District 96 | Matt Vitale (D) vs Brad Christ (R) | ||
| House District 101 | Paul Lentz (D) vs Ben Keathley (R) | ||
| House District 16 | Dan Kinate (D) vs Chris Brown (R) | ||
| House District 94 | Cathy A Summers (D) vs Maryellen Murphy (R) | ||
| House District 105 | Cat Clayton (D) vs Colin Wellenkamp (R) | ||
| House District 106 | Jewel Kelly (D) vs Travis Wilson (R) | ||
| House District 56 | Staci Czar (D) vs Steve Baker (R) | ||
| House District 104 | Tara Murray (D) vs Terri Violet (R) | ||
| House District 133 | Katelyn Zach (D) vs Jerry Angelo (R) | ||
| House District 31 | Justin Geiser (D) vs Ron Fowler (R) | ||
| House District 17 | Susan Hoisington (D) vs Bill Allen (R) | ||
| House District 30 | Kevin Grover (D) vs Jeff Martin (R) | ||
| House District 100 | Mary Ann Perkins (D) vs Philip Oehlerking (R) | ||
| House District 12 | Tracey Steele (D) vs Jerry Hickey (R) | ||
| House District 10 | LaTonya T. Williams (D) vs Paul Gatewood (R) | ||
| House District 89 | Bryan Troop (D) vs George Hruza (R) | ||
| House District 103 | Don Crozier (D) vs Dave Hinman (R) | ||
| House District 34 | Kemp Strickler (D) vs Sean Smith (R) | ||
| House District 95 | Meredith Sumenek (D) vs Bob Mahacek (R) | ||
| House District 20 | Jeff Spicer (D) vs Mike Steinmeyer (R) | ||
| House District 69 | Chris Chapman (D) vs Scott Miller (R) |
Senate seats in doubt
| District | Candidates | Chance before returns | Chance now |
|---|---|---|---|
| Senate District 8 | Keri Ingle (D) vs Jon Patterson (R) | ||
| Senate District 30 | Betsy Fogle (D) vs Melanie Stinnett (R) | ||
| Senate District 34 | Shereka Barnes (D) vs Mike Jones (R) |
Each counted figure names its basis: precincts is precincts reporting, est. vote is votes counted over votes expected, votes is the absentee batch plus the precincts reporting since (an estimate of the vote counted), reported is the share the election authority gave, final count is a hand-flagged complete count, and unknown means the source did not say. Seats in doubt: neither party has more than a 95% live chance, closest first. Chance before returns is the last pre-election forecast’s. Live seat chances move with the State Auditor swing, inside MO-2 with the US House race, and with any district totals entered from the Secretary of State.
How the live model works
Each county’s count against what it was expected to give
As each county reports, its count is compared with what it was expected to give, and every figure on this page is recalculated: the two chambers, the five ballot measures, the State Auditor race and MO-2. The model is rerun as the Secretary of State posts new county totals, and this page picks up each run on its own.
Before any returns, every county has an expected share for each contest: the statewide forecast spread across the counties by their past votes. As a county reports, the gap between its count and that expectation informs three things at once: a statewide swing, a swing for its region, and an urban-rural tilt, a swing that grows with how far the county leans from the state.
A county still counting is projected from its own partial count, shrunk toward the model, because the early batches (mail and early votes) differ from the final count; the less of it counted, the harder the shrink. A county that has not reported gets the swing plus its own uncertainty. The projections are added up by expected turnout in thousands of simulations, which give the range and the chances.
The State Auditor race is the only statewide partisan contest on the ballot, so its swing from expectation is carried into the seat model: the legislative chances move with it, and with any district totals entered from the Secretary of State.
The model was calibrated by replaying Missouri’s 2016-2024 county returns through simulated counting nights: 7 contests, 30 nights each. Its 80% intervals covered 73%-88% of outcomes, and the eventual winner was favoured at every checkpoint.
The replay, checkpoint by checkpoint
| Vote counted | Average miss, pts | Bias, pts | 80% intervals covering | 50% intervals covering | Winner favoured |
|---|---|---|---|---|---|
| 10% | 1.0 | −0.1 | 85% | 56% | 100% |
| 25% | 0.7 | 0.0 | 87% | 48% | 100% |
| 50% | 0.6 | −0.2 | 84% | 28% | 100% |
| 75% | 0.4 | −0.2 | 73% | 36% | 100% |
| 90% | 0.2 | −0.1 | 88% | 55% | 100% |
Source: data/forecast/nowcast_replay.json (model/nowcast_replay.py). Each night draws a pre-election error, a reporting order and an early-vote lean in metro counties; the truth is the official county result. Average miss and bias are the projected final share less the actual, in percentage points. Cases: pres 2016 -> 2024; governor 2016 -> 2024; US Senate 2016 -> 2024; auditor 2018 -> secretary of state 2024; abortion 2024, shaped by the presidential vote; min. wage 2024, shaped by the presidential vote; 2024 Amendment 7 (citizen-only voting), shaped by the presidential vote.
Sources5ShowHide
- Missouri Secretary of State, election-night results, November 3, 2026 (unofficial county returns) , accessed Nov 3, 2026.
- Missouri Secretary of State, official election results (grand totals, 2016 to 2024) , accessed Nov 1, 2026.
- Missouri Secretary of State, 2026 ballot measures (official titles, summaries and fair ballot language) , accessed Nov 1, 2026.
- Missouri Secretary of State, 2026 general election certified candidate list , accessed Nov 1, 2026.
- 2024 district leans: Secretary of State precinct results for president, governor and US Senate, allocated to the 2022 House and Senate district plans (Missouri Office of Administration) , accessed Nov 1, 2026.