Will Mary Peltola win the Alaska Senate race in 2026?

TL;DR

Mary Peltola’s forecast probability of winning is 84.6%, with 15.4% for No. The case for her rests on a narrow preliminary lead and a favorable historical late-count pattern. Treat her as the favorite, not the confirmed winner: outstanding ballots and ranked-choice transfers remain decisive.

Context

As of November 4, 2026, voting has ended but counting has not. The latest corroborated first-choice snapshot gives Peltola 102,701 votes and incumbent Dan S. Sullivan 102,075, a difference of 626 votes. Two other candidates hold votes that can transfer or exhaust during ranked-choice tabulation (GovGalaxy). The remaining uncertainty concerns ballots already cast, their validity and their rankings—not further campaigning.

The market resolves on the eventual election winner, including any run-offs. It requires agreement among the Associated Press, Fox News and NBC, or official certification if those calls do not agree. If results are not confirmed by July 31, 2027, the stated outcome is “Other.” December 1 is not a hard confirmation deadline under these criteria.

Evidence

The historical backbone is Peltola’s late-count performance in her previous Alaska general elections. Both the 2022 and 2024 House contests moved in her direction after the initial returns. These are useful comparisons because they involve the same candidate and the same state’s counting process. They are still only two elections, with different opponents and ballot compositions.

The following comparison keeps first-choice votes separate from ranked-choice transfers. The 2022 opening figures come from the November 9 report; the final figures are corroborated by federal election records. The 2024 opening figures come from a November 6 snapshot, followed by the November 30 final summary (2022 opening report, 2022 final results, 2024 early reporting, 2024 final summary).

First-choice measure	2022 House	2024 House
Peltola votes in opening snapshot	100,538	113,612
Peltola votes in final summary	128,553	152,828
Additional Peltola votes	28,015	39,216
Additional valid first choices, all candidates	50,281	78,609
Peltola share of additions	55.7%	49.9%

The 2024 comparison is especially useful because it was the less favorable of the two late counts. Begich’s first-choice lead declined from 11,020 votes to 6,722, a net Peltola gain of 4,298 votes. She improved without winning that race. The lesson is that late ballots can help her materially; they do not guarantee victory (Alaska’s News Source).

Snapshot timing changes the measured size of the shift. A later November 6 report showed Peltola at 114,511 and Begich at 124,969, with 252,310 total first choices. Against the final totals of 152,828, 159,550 and 329,555, respectively, that leaves a smaller subsequent Peltola net gain of 3,736 votes. These are different starting points, not conflicting final results (later opening report, final summary).

The older 2020 Senate comparison also showed a Democratic improvement during counting, but it is a weaker guide. Early snapshots used different counted totals and rounded shares, while pandemic-era voting produced an unusually large late-counted share. It supports caution about reading election-night margins as final; it does not justify carrying a fixed swing into this race (2020 Alaska Senate results).

Against that background, the current count leaves Peltola well positioned. GovGalaxy attributes its figures to the state’s November 4, 11:43 UTC file. It lists 102,701 votes for Peltola, 102,075 for the incumbent, 4,821 for Gerald Heikes and 3,583 for Daniel J. Sullivan Jr. The four candidates total 213,180 votes (GovGalaxy). Alaska Beacon separately reports the same 626-vote lead and identifies outstanding early, absentee and rural votes, with remaining rural votes believed to favor Peltola (Alaska Beacon).

That agreement supports the lead, but not every reporting detail. Beacon gives 213,873 ballots and 394 of 401 precincts; GovGalaxy gives 417 of 443 reporting units. The definitions remain unreconciled. Neither fraction should be read as the percentage of final ballots counted (Beacon, GovGalaxy).

Alaska’s counting process explains why this matters. Alaska Public Media reports that election-night returns historically cover about 70% of the eventual vote and that shifts of roughly four percentage points have occurred. Early returns already contain some early and absentee votes, so “late vote” is not one clean voting category (Alaska Public Media). Likewise, the reported 105,419 non-Election-Day ballots received is not an outstanding-ballot inventory: some were already counted (Alaska’s News Source).

Ranked-choice transfers are the main counterweight. The two minor candidates already hold 8,404 first-choice votes, far more than the leaders’ separation (GovGalaxy). But those ballots cannot simply be assigned to the incumbent. Some voters rank Peltola; others rank neither finalist, causing their ballots to exhaust.

The survey evidence explored both first choices and simulated final rounds. Its full sequence is mixed: most surveys put Peltola ahead at the end, while the latest InsiderAdvantage poll showed a tie. These are voter surveys, not observed transfers.

Survey and fieldwork	Initial choices: Peltola / incumbent / other Sullivan / Heikes	Simulated final
AARP, September 8–11	46% / 41% / 4% / 3%	53% / 47%
ASR, October 1–3	47% / 45% / 5% / 3%	51% / 49%
Data for Progress, October 8–16	47% / 44% / 4% / 1%	51% / 49%
InsiderAdvantage, October 9–10	48% / 43% / 2% / 2%	50% / 48%, with 2% undecided
ASR, October 26–29	47.7% / 45.9% / 3.4% / 3%	50.8% / 49.2%
InsiderAdvantage, October 31–November 1	48% / 48%; minor-candidate figures not reported here	50% / 50%

The ASR decimals came from secondary reporting, and minor-candidate subsamples across these surveys were small. The sound inference is limited: Sullivan can gain through transfers, but overwhelming consolidation is not established. Alaska Public Media also reports that minor-candidate supporters rank more often than supporters of finalists, so statewide bullet-voting rates are not valid transfer estimates for this contest (Alaska Public Media).

What's non-obvious

The important distinction is between a tiny lead and a favorable position. Peltola’s 626-vote advantage is not secure on its own. Her stronger position comes from holding that lead before counting ballots that historical comparisons and local reporting suggest will improve her margin. A close first-choice count therefore does not automatically imply an evenly balanced eventual result.

The opposite shortcut also fails: a favorable late count does not eliminate transfer risk. First-choice additions and ranked-choice redistribution are separate processes. Peltola can gain during one and lose ground during the other. The forecast favors her because the late-count evidence points her way, while retaining the possibility that a less favorable remainder and stronger Sullivan transfers reverse the outcome.

Uncertainties

The largest gap is a verified inventory of accepted but uncounted ballots, broken down by geography and voting method. Without it, the favorable late-count expectation rests on local reporting and a small historical reference class. A channel-level inventory and comparable historical results would show whether this year’s remainder resembles the earlier elections.

Actual rankings are also unknown. No verified current transfer or exhaustion figures establish how Heikes and other-Sullivan voters divide. Cast-vote records or the official ranked-choice tabulation would close that gap. Ranked-choice results are scheduled for November 18, which is also the qualifying overseas-ballot receipt deadline reported by AP (AP).

Current reporting has a further unresolved discrepancy. Fox showed a smaller count with Sullivan ahead and left the race uncalled. That is consistent with an older return snapshot, but a directly verified current official tally is needed to settle the difference (Fox News). The available evidence also does not establish the required unanimous media call.

Finally, litigation remains a residual confirmation risk. AP reports that the suit seeking to block certification was dismissed and appealed. That does not establish that the result will be overturned, but it leaves a legal issue outstanding (AP). The central uncertainty remains the ballots and their rankings, not ordinary counting delay.