A September poll showed Lauren Boebert leading Eileen Laubacher 45% to 40% in Colorado's CO-04, a narrow edge that still leaves the race unsettled with a little over a day left before voting finishes. Kalshi, meanwhile, gives Boebert a 64% probability of retaining her seat, putting prediction-market confidence ahead of the polling margin.
That split matters because Colorado has only four congressional races that are at least somewhat competitive, and public polling has been sorely lacking across the state. In that vacuum, each new number gets outsized attention, especially when it comes from a contest that could help shape the 2026 midterm election map.
CO-04 covers the Eastern Plains and some of the Front Range, which makes it one of the districts where any shift in turnout or late movement could matter. The September survey was sponsored by an anti-Boebert PAC, but even that snapshot was not enough to erase the broader picture Kalshi is pricing in: Boebert is still favored, just not by the same margin the poll suggests.
The gap is easy to explain. A 45% to 40% lead in a single poll shows where the race stood when voters were surveyed; a 64% winning probability reflects how a market weighs that result against the chance that other unpublished polls, turnout patterns, or late changes in opinion could still move the outcome. In a state where prediction markets are the only published numbers for many contests, that distinction can matter more than the headline lead itself.
What comes next is simple enough. With voting nearing its end, the last public signals before Election Day are likely to be treated as a guide rather than a verdict, and Boebert's lead in one September poll will be judged against whether the final stretch narrows the race or confirms her hold on CO-04.
