Democrats are slightly favored in the Senate.

55in 100 Democratic

Forecast as of October 27, 2026: Democrats have a 55 in 100 chance to control the Senate, Republicans 45 in 100. Current race-by-race calls: D 52 - R 48.

Senate control and seat distribution

Expected D seats: 51.0**includes 2.5 independents expected to caucus with the Democrats (2 sitting, 0.5 from this year’s races)
Expected R seats: 49.0
Current Senate: D 47 · R 5345 D + 2 I aligned
Change: D+4

5 in 100: the majority hinges on independent winners — neither party reaches its threshold without them.

Lower 10% tailMiddle 80% · 46–56 DUpper 10% tail

D 51 · R 49The median outcome — 10.6% of simulations land exactly here: Democratic control, with Democratic-aligned +4 and Republicans −4 from today's Senate.55.3% land here or better for Democratic-aligned — 44.7% land worse.

From today’s Senate Today · Selected

100,000 joint simulations · each outcome totals 100 seats.

How the forecast has changed

Oct 27 · latestDemocrats 55 in 100Republicans 45 in 100seats 51.0 D to 49.0 R

Reconstructed with the current model Aug 1 – Oct 5; published readings Oct 6 – Oct 27.

Current 80% range · D 46–56 · R 44–54. Democratic seats include 2.5 independents expected to caucus with the Democrats (2 sitting, 0.5 from this year's races), as in the figure above. Sep 16 – Sep 18 published no Republican seat figure; it is 100 minus the Democratic one, which holds exactly on the days either side.

Data table

Date	D control	R control	D seats	R seats	Reading
Oct 27	55 in 100	45 in 100	51.0	49.0	Published forecast
Oct 26	56 in 100	44 in 100	51.0	49.0	Published forecast
Oct 25	55 in 100	45 in 100	51.0	49.0	Published forecast
Oct 24	55 in 100	45 in 100	51.0	49.0	Published forecast
Oct 23	55 in 100	45 in 100	50.9	49.1	Published forecast
Oct 22	56 in 100	44 in 100	51.0	49.0	Published forecast
Oct 21	56 in 100	44 in 100	51.0	49.0	Published forecast
Oct 20	56 in 100	44 in 100	51.0	49.0	Published forecast
Oct 19	56 in 100	44 in 100	51.1	48.9	Published forecast
Oct 18	56 in 100	44 in 100	51.1	48.9	Published forecast
Oct 17	56 in 100	44 in 100	51.1	48.9	Published forecast
Oct 16	56 in 100	44 in 100	51.0	49.0	Published forecast
Oct 15	55 in 100	45 in 100	51.0	49.0	Published forecast
Oct 14	55 in 100	45 in 100	51.0	49.0	Published forecast
Oct 13	55 in 100	45 in 100	51.0	49.0	Published forecast
Oct 12	54 in 100	46 in 100	50.9	49.1	Published forecast
Oct 11	54 in 100	46 in 100	50.9	49.1	Published forecast
Oct 10	54 in 100	46 in 100	50.9	49.1	Published forecast
Oct 9	54 in 100	46 in 100	50.8	49.2	Published forecast
Oct 8	54 in 100	46 in 100	50.9	49.1	Published forecast
Oct 7	54 in 100	46 in 100	50.9	49.1	Published forecast
Oct 6	52 in 100	48 in 100	50.7	49.3	Published forecast
Oct 5	52 in 100	48 in 100	50.7	49.3	Reconstructed with the current model
Oct 4	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Oct 3	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Oct 2	50 in 100	50 in 100	50.5	49.5	Reconstructed with the current model
Oct 1	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Sep 30	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Sep 29	52 in 100	48 in 100	50.6	49.4	Reconstructed with the current model
Sep 28	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Sep 27	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Sep 26	51 in 100	49 in 100	50.6	49.4	Reconstructed with the current model
Sep 25	50 in 100	50 in 100	50.5	49.5	Reconstructed with the current model
Sep 24	49 in 100	51 in 100	50.5	49.5	Reconstructed with the current model
Sep 23	49 in 100	51 in 100	50.5	49.5	Reconstructed with the current model
Sep 22	49 in 100	51 in 100	50.4	49.6	Reconstructed with the current model
Sep 21	47 in 100	53 in 100	50.3	49.7	Reconstructed with the current model
Sep 20	48 in 100	52 in 100	50.3	49.7	Reconstructed with the current model
Sep 19	48 in 100	52 in 100	50.3	49.7	Reconstructed with the current model
Sep 18	48 in 100	52 in 100	50.3	49.7	Computed after the fact; no edition was issued that day (Republican seats are 100 minus the Democratic figure)
Sep 17	46 in 100	54 in 100	50.2	49.8	Computed after the fact; no edition was issued that day (Republican seats are 100 minus the Democratic figure)
Sep 16	47 in 100	53 in 100	50.2	49.8	Computed after the fact; no edition was issued that day (Republican seats are 100 minus the Democratic figure)
Sep 15	49 in 100	51 in 100	50.4	49.6	Reconstructed with the current model
Sep 14	49 in 100	51 in 100	50.4	49.6	Reconstructed with the current model
Sep 13	47 in 100	53 in 100	50.2	49.8	Reconstructed with the current model
Sep 12	46 in 100	54 in 100	50.1	49.9	Reconstructed with the current model
Sep 11	46 in 100	54 in 100	50.1	49.9	Reconstructed with the current model
Sep 10	46 in 100	54 in 100	50.1	49.9	Reconstructed with the current model
Sep 9	43 in 100	57 in 100	49.9	50.1	Reconstructed with the current model
Sep 8	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Sep 7	43 in 100	57 in 100	49.9	50.1	Reconstructed with the current model
Sep 6	43 in 100	57 in 100	49.9	50.1	Reconstructed with the current model
Sep 5	44 in 100	56 in 100	50.0	50.0	Reconstructed with the current model
Sep 4	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Sep 3	43 in 100	57 in 100	49.9	50.1	Reconstructed with the current model
Sep 2	43 in 100	57 in 100	49.9	50.1	Reconstructed with the current model
Sep 1	44 in 100	56 in 100	50.0	50.0	Reconstructed with the current model
Aug 31	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Aug 30	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Aug 29	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Aug 28	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Aug 27	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Aug 26	43 in 100	57 in 100	49.9	50.1	Reconstructed with the current model
Aug 25	43 in 100	57 in 100	50.0	50.0	Reconstructed with the current model
Aug 24	42 in 100	58 in 100	49.9	50.1	Reconstructed with the current model
Aug 23	42 in 100	58 in 100	49.9	50.1	Reconstructed with the current model
Aug 22	42 in 100	58 in 100	49.9	50.1	Reconstructed with the current model
Aug 21	40 in 100	60 in 100	49.8	50.2	Reconstructed with the current model
Aug 20	40 in 100	60 in 100	49.8	50.2	Reconstructed with the current model
Aug 19	40 in 100	60 in 100	49.8	50.2	Reconstructed with the current model
Aug 18	39 in 100	61 in 100	49.7	50.3	Reconstructed with the current model
Aug 17	39 in 100	61 in 100	49.7	50.3	Reconstructed with the current model
Aug 16	38 in 100	62 in 100	49.7	50.3	Reconstructed with the current model
Aug 15	39 in 100	61 in 100	49.7	50.3	Reconstructed with the current model
Aug 14	38 in 100	62 in 100	49.7	50.3	Reconstructed with the current model
Aug 13	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 12	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 11	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 10	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 9	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 8	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 7	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 6	36 in 100	64 in 100	49.5	50.5	Reconstructed with the current model
Aug 5	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 4	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 3	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 2	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model
Aug 1	37 in 100	63 in 100	49.6	50.4	Reconstructed with the current model

Each reconstructed day is the current model on that day’s own inputs, not a forecast published then. Corrections are dated in the changelog.

All 35 Senate races

NC GA ME MI NH OH TX IA NE AK MN IL FL CO NM NJ VA MA RI DE OR AL AR ID KS KY LA MS MT OK SC SD TN WV WYVTWANDWINYNVINPACTCAUTMOMDAZHI

35 Senate races, one person covering them. Even a filibuster needs 41. Get what moved by email Support the work

The Flippers

3 of 35 seats are projected to change hands. Not counted: 2 toss-ups, Alaska and Maine, whose slight favorites would also change the party. Hover or tap for the forecast.

The modeled races · closest first

Sort8 of 35

MaineTroy Jackson (D) vs. Susan Collins (R)Toss-upD+0.3 TexasFlipJames Talarico (D) vs. Ken Paxton (R) · openTilt DD+0.7 AlaskaMary Peltola (D) vs. Dan S. Sullivan (R) vs. Gerald L. Heikes (R)Toss-upD+0.9 IowaJosh Turek (D) vs. Ashley Hinson (R) · openTilt RR+1.0 MichiganAbdul El-Sayed (D) vs. Mike Rogers (R) · openTilt DD+1.3 NebraskaPete Ricketts (R) vs. Dan Osborn (I) · multiparty fieldTilt RR+1.4 OhioFlipSherrod Brown (D) vs. Jon Husted (R)Tilt DD+1.8 KansasAdam Hamilton (D) vs. Roger Marshall (R)Tilt RR+1.9

See every competitive race on The Map →

Which races decide control

Win it or lose it. The Democrats’ chance of the Senate on the nights they win each race, and on the nights they lose it.

MinnesotaFlanagan (D) 64 in 100Flanagan (D) wins Minnesota in 63,746 of the 100,000 simulations; in those, Democrats win the Senate 73 times in 100. In the other 36,254, 24 times in 100.
OhioBrown (D) 63 in 100Brown (D) wins Ohio in 62,607 of the 100,000 simulations; in those, Democrats win the Senate 74 times in 100. In the other 37,393, 25 times in 100.
MichiganEl-Sayed (D) 59 in 100El-Sayed (D) wins Michigan in 59,027 of the 100,000 simulations; in those, Democrats win the Senate 75 times in 100. In the other 40,973, 27 times in 100.
GeorgiaOssoff (D) 92 in 100Ossoff (D) wins Georgia in 91,248 of the 100,000 simulations; in those, Democrats win the Senate 60 times in 100. In the other 8,752, 12 times in 100.
IowaTurek (D) 43 in 100Turek (D) wins Iowa in 43,245 of the 100,000 simulations; in those, Democrats win the Senate 82 times in 100. In the other 56,755, 35 times in 100.
KansasHamilton (D) 37 in 100Hamilton (D) wins Kansas in 37,008 of the 100,000 simulations; in those, Democrats win the Senate 84 times in 100. In the other 62,992, 38 times in 100.
New HampshirePappas (D) 81 in 100Pappas (D) wins New Hampshire in 80,581 of the 100,000 simulations; in those, Democrats win the Senate 64 times in 100. In the other 19,419, 18 times in 100.
NebraskaOsborn (I) 41 in 100Osborn (I) wins Nebraska in 41,331 of the 100,000 simulations; in those, Democrats win the Senate 82 times in 100. In the other 58,669, 37 times in 100.
TexasTalarico (D) 55 in 100Talarico (D) wins Texas in 55,189 of the 100,000 simulations; in those, Democrats win the Senate 75 times in 100. In the other 44,811, 31 times in 100.
MaineJackson (D) 52 in 100Jackson (D) wins Maine in 52,140 of the 100,000 simulations; in those, Democrats win the Senate 76 times in 100. In the other 47,860, 33 times in 100.
AlaskaPeltola (D) 51 in 100Peltola (D) wins Alaska in 51,309 of the 100,000 simulations; in those, Democrats win the Senate 76 times in 100. In the other 48,691, 33 times in 100.
South CarolinaAndrews (D) 26 in 100Andrews (D) wins South Carolina in 26,432 of the 100,000 simulations; in those, Democrats win the Senate 84 times in 100. In the other 73,568, 45 times in 100.

Races move together: a night Democrats carry one of these is usually a good night for them everywhere. A row's gap mostly shows what kind of night it is, not what that one seat is worth.

Each figure is a count over the same 100,000 simulations as the Senate figure at the top of the page. The smallest group behind any figure here is 8,752 simulations. 12 of the 35 races are listed: the ones most often the deciding seat, among the 18 that each side wins in at least 5,000 simulations. The other 17 are too one-sided to split this way.

The chance beside each name is that race's own forecast, as its page prints it. The simulations are drawn for the whole chamber at once, and each race listed here is won in them within 0.6 points of that chance: New Hampshire is won in 80,581 of 100,000, against 81 in 100.

Pick the winners yourself in the Scenario Machine

What if the national environment moved?

Explore Senate and House scenarios

Where we disagree with the markets

Democratic win chance · forecast vs. Polymarket. Markets never feed the forecast. Quotes: Oct 27, 2026.

Race
Alaska	51 in 100	74 in 100	23 pts more Democratic in market
Minnesota	64 in 100	86 in 100	22 pts more Democratic in market
Texas	55 in 100	67 in 100	12 pts more Democratic in market
New Hampshire	81 in 100	92 in 100	11 pts more Democratic in market
Maine	52 in 100	62 in 100	10 pts more Democratic in market

Biggest movessince Oct 25, 8:58 PM ET

South DakotaBrian Bengs (I) v. Mike Rounds (R)+1.1R+15.2
South Carolina+0.8R+4.1
New Hampshire-0.3D+4.9
Iowa+0.2R+1.0

← Toward Republicans±5 pointsToward Democrats →

Green is an independent candidate, not a party: South Dakota, Brian Bengs (I) v. Mike Rounds (R). The bar points toward whichever of the two gained; the independent's side of the axis is the one opposite the major-party candidate.

4 of 35 Senate races moved at least 0.2 points · Mississippi, Montana not compared on this board