POLLS · QUEBEC
Quebec polls
Provincial polls published since the October 5, 2026 election, compared with the actual vote and translated into seats. We add each new poll as soon as it is published.
1 poll since the election · updated October 22, 2026
Latest poll: Léger
| Party | Poll | Oct. 5 vote | Change (points) | Reading |
|---|---|---|---|---|
| PQ | 30% | 28.0% | +2.0 | within margin |
| PLQ | 26% | 24.2% | +1.8 | within margin |
| PCQ | 22% | 20.7% | +1.3 | within margin |
| QS | 12% | 12.4% | −0.4 | within margin |
| CAQ | 8% | 13.4% | −5.4 | down |
What this poll would mean in seats
PQ 59 · PLQ 41 · PCQ 19 · QS 8 · majority : 64
2 ridings would change hands: Sherbrooke (QS → PLQ), Vachon (PQ → PLQ).
In brief
- Changes larger than the margin of error: the CAQ −5.4.
- In seats, the PQ would get about 59 seats, short of the 64-seat majority (the PQ won 60 on October 5).
- Caution: these are not regular voting intentions, but a hypothetical question asked after the election.
In the same poll, the reported October 5 vote (PQ 28, PLQ 24, PCQ 21, CAQ 13, QS 12) is very close to the official result.
How the pollsters did in 2026
Each firm’s average error on the five main parties, for its last poll before the October 5 vote. The smaller the number, the closer the firm was to the result.
| Firm | Fieldwork | Average error |
|---|---|---|
| Forum Research | Oct. 3–4 | 0.81 pt |
| Léger | Sept. 30–Oct. 1 | 0.90 pt |
| Pallas Data | Oct. 3 | 0.92 pt |
| Research Co. | Oct. 2–4 | 1.44 pt |
| Liaison Strategies | Oct. 2–3 | 1.80 pt |
| Synopsis Recherche | Sept. 24–27 | 2.46 pt |
| Mainstreet Research | Oct. 2–3 | 2.84 pt |
| Nanos Research | Sept. 12–14 | 3.39 pt |
| Segma Recherche | Sept. 8–17 | 3.59 pt |
Polls conducted earlier in the campaign could not capture late shifts: part of the gap comes from timing, not only the firm. Our full analysis →
How we turn a poll into seats
We start from the official October 5, 2026 results in each of the 127 ridings (Vachon after the judicial recount). If a party goes from 28% to 30% province-wide, its share is multiplied by 30/28 in every riding, then shares are rescaled to 100%. The leading party wins the seat.
This is a simple estimate, not the full projection 127QC publishes during campaigns: it ignores regional shifts, candidates and uncertainty. It answers one question: what would the Assembly look like if the vote moved the same way everywhere?
The margin of error shown is that of a random sample of the same size, at the 95% confidence level. For online polls, firms give it for comparison purposes.
Further reading
127QC is independent and ad-free. Support the project