Saint-Hyacinthe—Bagot—Acton — Election Outlook & Projection
Where this riding stands
The RidingWatch model currently projects the Bloc Québécois at 43.6% in Saint-Hyacinthe—Bagot—Acton, 7 points ahead of the Liberals at 36.6%, with the Conservatives at 10.9%, the NDP at 5.9%, the Greens at 3%. Across 10,000 simulations the seat goes to the Bloc Québécois in 90% of them, the Liberals in 10%. The simulations put the Bloc Québécois’ share anywhere from 39% to 49%, and the Liberals' from 31% to 43%.
In April 2025 the Bloc Québécois won here with 44%, 10 points ahead of the Liberals. The Bloc Québécois also held it in 2021. Compared with the 2025 result, the current projection has the Conservatives down 7.1, the NDP up 3.9, the Liberals up 2.6 points. These shifts come from the national polling swing applied to this riding's past results — not from riding-level polling, which the model does not have.
The seat is currently vacant — Simon-Pierre Savard-Tremblay resigned June 19, 2026. A by-election will fill it unless a general election is called first.
Rating: Likely Bloc Québécois — the Bloc Québécois is ahead in most simulations. In practical terms, the Bloc Québécois are ahead in most simulations, but the Liberals win in 10% of them — the outcome is probable, not fixed. Within 10 points of the lead: the Liberals (36.6%). Every vote counts toward a party's national total regardless of the local outcome. The Strategic Voting Assistant below shows how these standings look for a specific goal, including simply voting for your first choice.
Sitting member
About this riding
Located in the Montérégie region, this riding encompasses the city of Saint-Hyacinthe and surrounding communities. Saint-Hyacinthe serves as a significant regional manufacturing and agricultural centre with diverse industries. Surrounding areas feature small towns, farms, and transitional terrain. Agricultural processing and small-scale manufacturing provide economic activity. The riding combines urban centre functions with rural agricultural character. Regional service and institutional facilities serve broader populations alongside local communities.
Projected Vote Share — Next Federal Election
Projected using a blended proportional-uniform swing model with incumbency adjustments. Based on the current national polling average and regional crosstabs as of 2026-09-14.
Vote-share trend
2021 and 2025 are official results for this riding; the third point is the current model projection. No estimated historical points are shown.
2025 Election Results vs. Current Projection
| Party | 2025 Vote % | Projected % | Change |
|---|---|---|---|
| Liberal | 34% | 36.6% | +2.6 |
| Conservative | 18% | 10.9% | -7.1 |
| NDP | 2% | 5.9% | +3.9 |
| Bloc Québécois | 44% | 43.6% | -0.4 |
| Green | 1.4% | 3% | +1.6 |
Strategic Voting Assistant
Reasoning uses the current projected shares for this riding. Viability = within 10 points of the leader or at least 15%.
Explore More Ridings
Longueuil—Saint-Hubert Berthier—Maskinongé Rimouski—La Matapédia Laurentides—Labelle Les Pays-d’en-Haut
About the Model
This projection is generated by the RidingWatch statistical model, which combines national polling data with historical election results to estimate vote shares and likely outcomes in each of Canada's 343 federal ridings.
The model begins with a weighted blend of results from the 2025 and 2021 federal elections, then applies the polling swing to each riding using a mostly-uniform (90/10 uniform/proportional) swing method — the national swing, adjusted by the province's pooled deviation where pollsters publish regional crosstabs. No pollster house effects are applied. Incumbency is not assumed: only first-term MPs (+1) and party leaders (+2) receive an adjustment. Uncertainty is estimated through 10,000 Monte Carlo simulations with correlated regional errors.
Projections are updated as new polling data becomes available. All estimates are probabilistic — they represent the most likely outcome given current information, not a guarantee of any result. Actual election outcomes depend on factors the model cannot fully capture, including local campaigns, turnout variation, and late-breaking events.