Fox News poll finds Texas Senate close at 51-49 as our forecast edges to 56-44
The new survey nudged our model's Democratic chance up one point and our projected margin by a tenth, a change too small to alter the picture of a close race.
Written by the newsroom · anthropic/claude-opus-5.5 · October 1, 2026 at 7:02 PM EDT
Correction: the forecasts in this article predate a polling-data repair. Some imported polls omitted candidates who shared a party, and ranked-choice rounds were not consistently selected. The current forecasts use corrected results and recalculated pollster adjustments. See current forecasts.
A Beacon Research/Shaw & Company Research survey for Fox News, fielded September 24-28 among 881 likely voters, has James Talarico, the Democrat, at 51% and Ken Paxton, the Republican, at 49%. A two-point gap in a poll of that size is not a meaningful lead. The survey describes a close race.
Our model now gives Talarico a 56 in 100 chance of winning the open seat, up from 55% in the run completed earlier this evening. Paxton's chance is 44%, down from 45%. Our projected margin moved from D+0.7 to D+0.8.
The poll confirmed more than it changed. It sits close to where our model already had the race, so it pulled the estimate only slightly. Our range of plausible outcomes remains wide: the 5th to 95th percentile now runs from R+8.4 to D+10.0, against R+8.5 to D+9.8 before. Either candidate winning by several points falls well inside that band.
The seat matters for the national picture. Our model gives Democrats a 53% chance of winning Senate control and Republicans 46%, with Democrats averaging 50.6 seats in our simulations. Democrats need 51 seats for control; Republicans need 50, because the vice president is a Republican and a 50-50 tie goes to them. With 33 days to the election, our model treats Texas as a race either party could win.
Polls this dispatch was written from
- Beacon Research/Shaw & Company Research · Sep 28, 2026 · n=881 · Texas Senate
Every constant behind these numbers is on the methodology page.