The latest multi-level regression and post-stratification (MRP) modelling, based on polling data collected through April and early May 2026, projects that if a general election were held today, Labour would win approximately 290–320 seats — losing its current majority but remaining the largest party. Reform UK, despite polling at 26% nationally, could win between 60 and 90 seats, a figure that would make it the second-largest party in the Commons by seats for the first time.
What Is MRP and Why Does It Matter?
MRP (Multi-level Regression and Post-stratification) is a polling methodology that uses large samples — typically 10,000 to 50,000 respondents — combined with detailed demographic and geographic data to produce constituency-level vote estimates. Unlike uniform national swing (UNS) models, which assume the same shift applies everywhere, MRP attempts to model how different types of people in different types of places will vote.
The technique was popularised in UK politics by YouGov’s 2017 and 2019 MRP polls, which accurately identified the 2017 hung parliament outcome when other methods failed. In 2024, YouGov’s MRP was the closest major projection to the final result. Several organisations now produce regular MRP estimates, including Electoral Calculus, Focaldata, and a proprietary model run by Sky News in partnership with a major polling firm.
The Current Projection: Seat by Seat
Averaging across three MRP models that have been published in 2026 (Electoral Calculus April, Focaldata March, and a model published by a Sunday broadsheet in February), the central projection is approximately: Labour 290–320 seats, Conservatives 140–165 seats, Reform UK 60–90 seats, Liberal Democrats 70–90 seats, SNP 15–25 seats, others 20–30 seats. These ranges reflect different assumptions about the uniformity of swing, tactical voting, and turnout effects.
The 326-seat threshold for a parliamentary majority means all current models show Labour losing its overall majority, though in some scenarios it would be just a few seats short. The most optimistic projection for Labour (based on the March 2026 Focaldata data) shows 324 seats — technically a majority of one. The most pessimistic (Electoral Calculus April) shows 288 seats, requiring Labour to seek coalition or confidence-and-supply support.
How FPTP Distorts Reform’s 26%
The most striking feature of current MRP projections is the disparity between Reform UK’s national vote share (26%) and its projected seat total (60–90). In GE2024, Reform won 14.3% of votes for five seats. Even at 26%, the party faces a structural disadvantage rooted in the geographic distribution of its support.
Reform’s vote is broadly distributed across England and Wales. It polls at 20–35% in most constituencies but rarely exceeds 40% in any single seat, the threshold typically needed to win under FPTP against divided opposition. By contrast, Labour’s 27% is highly concentrated in urban areas where it routinely polls 45–65% in safe seats, generating large margins that translate efficiently into seat totals.
The seats where Reform is projected to win under current conditions are predominantly former Conservative strongholds in the East of England, East Midlands, and coastal towns, where the Conservative vote has collapsed to single figures and Reform has emerged as the primary right-of-centre option. In some of these seats, Reform leads on 38–42% in MRP constituency estimates, making them winnable even under FPTP.
The D-Day Scenarios
MRP modellers typically produce scenario analyses alongside central projections. The key variables are: (1) whether Labour’s vote holds or continues to decline; (2) whether Conservative voters continue to split between the Tories and Reform, or begin to consolidate; (3) whether tactical voting against Reform materialises in key marginals.
In the “Reform surge” scenario (Reform at 30%, Conservatives at 18%), current MRP models project Reform winning 110–130 seats, potentially becoming the official opposition. In the “Conservative recovery” scenario (Tories at 28%, Reform at 22%), Reform is projected to win just 20–35 seats — illustrating how sensitive the seat projections are to the Conservative-Reform split.
The Tactical Voting Question
A critical unknown in all MRP projections is the degree to which anti-Reform tactical voting will materialise. In 2024, there was significant Labour-Liberal Democrat tactical voting against the Conservatives. If a similar dynamic emerges — with Labour and Lib Dem voters coordinating to keep Reform out in marginals — the seat projections would be significantly more favourable to Labour. Survey evidence on tactical voting intentions is inconclusive: 42% of Labour voters say they would consider voting tactically to keep Reform out, but only 28% say they would definitely do so. Track the latest MRP projections here.
Related: MRP projection tracker → • GE2029 hub → • What is MRP? →