Ohio Early Vote 2026

Loading…

Source: the Ohio Secretary of State's daily absentee file (every early in-person and mail ballot request), its weekly statewide voter file, its absentee dashboard (official daily returns by method for all 88 counties since the 2024 general), and the voter-level absentee exports that county boards publish. Counts are ballots returned by the as-of date.

Where 2026 stands

By party at the same day

2026 and 2024: SOS dashboard, all 88 counties (party as the counties report it). Panel columns: the county-export panel, with party rebuilt as of each election from primary history.

Cumulative ballots by day

Each line is the running total of returned ballots by days before Election Day. Ohio elections fall on a Tuesday, so the same day count lands on the same weekday each cycle. Past cycles run to their final count; 2026 stops at the as-of day. Past cycles appear only where a source has them (see comparison counties).

Show data tables for these charts

2026 vs 2022 and 2024 at the same day

Every count is cumulative through days before each Election Day. The 2024 comparison covers all 88 counties: it comes from the Secretary of State's absentee dashboard, which reports official daily returns by true method since the 2024 general. The 2022 midterm isn't in the dashboard. It's rebuilt from the voter-level exports that most county boards publish. Early in-person voting opened 28 days out in 2024 and 2026 but 27 days out in 2022, so 2022 has one fewer in-person day at any given point.

Statewide, all 88 counties (SOS dashboard)

Counties with 2022 history:

The 12 largest counties

Every county

Click a row to make that county the geography for every section.

2024 columns: SOS dashboard. *2022 columns: the county's own export, compared with that export's 2026 count. Counties without a 2022 export show –.

Cohort retention: who is coming back

Voters are matched across elections by their state voter ID (never published). Every 2026 figure is shown next to the same figure for the previous pair of elections spanning the same gap, at the same day before Election Day. Past ballot methods come from the county exports, so this runs on the counties whose exports cover all the elections involved: .

"Party then" is the voter's party as of the base election, rebuilt from primary history (Ohio party = last partisan primary ballot within the prior two calendar years). Both cohorts come from today's voter file, so voters removed since the base election are missing from both pairs.

Show data table

Who's turning out, per 100,000 registrants

Early ballots divided by registered voters in the same group, per 100,000 of them, so groups of different size can be compared. 2026 is set against 2022 at the same day. Party, vote-history group and turnout quartile are rebuilt as of each election from the voter file, so 2022 is grouped the way it looked in 2022. Approach adapted from NC Early Vote and Joel's Early Vote Tracker. Panel: .

By party and vote history

2026 groups: voted the 2022 midterm; presidential-only (2020 or 2024, not 2022); new since the 2020 general; lapsed (none of the three). The 2022 baseline uses the equivalent windows: 2018; 2016 or 2020; since 2016. The bar runs from 0.5× (left) to 2× (right) on a log scale, with the tick at 1×. "<10" means fewer than 10 ballots.

The race

Cumulative ballots per 100k registrants by day, Democrats against Republicans. 2026 is solid and stops at the as-of day; 2022 is dashed and 2024 dotted.

By turnout-propensity quartile

Every registrant gets a turnout score from past participation alone: 5 prior generals (same-type elections ×2) plus the 3 prior even-year spring primaries (×0.5), over the elections each voter was registered for. Quartiles are equal-sized among that cycle's panel registrants; Q1 is the least likely quarter to vote, Q4 the most. 2024 is a presidential year, shown for context.

Democratic share of Democratic + Republican ballots, by quartile

Unaffiliated and minor-party ballots are left out of this share.

Turnout by this year's primary ballot

Ohio's primaries are semi-open: any voter can take either party's ballot (which makes them that party's member) or an issues-only ballot. This groups registrants by the ballot they took in that year's May primary. Those who skipped it are grouped by the last partisan ballot they ever took. Turnout so far is compared with the same day in 2022. Panel: .

Turnout by day

Share of registrants who have voted, by May primary ballot

Who is voting: vote history and party switching

The 2026 electorate to date, split by what each voter did in the last comparable election, set beside the 2022 electorate on the same day. Panel: .

Party switching

Of early voters so far who also voted in the last midterm: the row is their party then, the columns their party now. In Ohio, switching means taking the other party's primary ballot, or skipping primaries long enough to become unaffiliated.

2026 voters so far, by party in 2022

2022 voters at the same day, by party in 2018

Map: where turnout is high and who is voting

Every county on one scale. Click or tap a county to make it the geography for every section.

Show map data as a table

Counties and districts

2026 against 2024 on the same day for every county (SOS dashboard); legislative districts show 2026. Party share is of all ballots in the area. Cells with fewer than 10 ballots show "<10", never zero. Click a row to select it.

County deep dive

How ballots came back (mail, drop box, in person, delivered to the board) comes from the Secretary of State's dashboard for every county. Ballot problems that need curing (missing ID, signature mismatch…) come from the counties that publish voter-level exports (★).

How ballots came back

Ballot problems flagged (2026)

Same-day comparison by party and method

"Same day" = cumulative through days before each Election Day. 2026 and 2024: SOS dashboard. *2022: the county's own export, compared with that export's 2026 count. †Final 2022 total from the county's voter file, which records each voter's ballot method; voters removed since are missing, so it's a floor.

Methodology and caveats