Quebec 2026: what the results really change

The October 5 vote profoundly reshapes political representation in Quebec. The Parti Québécois leads in 59 of the 127 ridings, ahead of the Quebec Liberal Party in 40, the Conservative Party of Quebec in 19 and Québec solidaire in nine. The Coalition Avenir Québec wins more than half a million votes but does not finish first in any riding. [S01]

The provincial result deserves to be read at two scales. Votes show how many ballots each party received. First places in the ridings determine its representation. In 2026, the gap between these two measures is particularly visible: the PQ wins about 28% of valid votes and leads in nearly 46.5% of ridings. In this table, no party reaches the arithmetic threshold of 64 out of 127. [S01; LQV calculations]

This analysis explains the main changes, the territorial differences and the questions that remain open. It does not claim to know the individual motivations of voters.

Data status. We use the Élections Québec feed updated on October 6 at 2:23 a.m., Quebec time. It covers all 20,704 polling stations and flags the provincial data and all 127 ridings as final. The expression “leading” describes the ranking in this file; it does not replace an official declaration of election, a court decision or an announcement on the formation of the government. The figures and text will have to be revised if the source publishes a correction. [S01; S04]

Six key facts

• The PQ comes first provincially in both votes and ridings, with 28.02% of the vote and 59 first places.
• The PLQ wins 24.22% of the vote and leads in 40 ridings.
• The PCQ wins 20.67% of the vote and 19 first places, compared with no seats in the 2022 general election.
• The CAQ wins 13.35% of the vote without a single first place; QS wins 12.43% and nine first places.
• Eight races end with fewer than 500 votes between the top two candidates. Ten have a gap of less than two percentage points.
• In 100 of 127 ridings, the leading candidate receives less than half of the valid votes. [S01; S02; LQV calculations]

1. The provincial picture: five parties, four leading locally

Source: Élections Québec. Descending order of votes, the same for every party. Other parties and independent candidates remain in the complete data and in the calculations; they are not removed from the denominator. [S01]

The first observation is how widely support is spread. The three parties with the most votes each exceed 20%, but none reaches 30%. That makes local results decisive: a lead of a few points over opponents who are themselves close can be enough to finish first.

The second observation concerns the CAQ. Its 571,520 votes represent a real electorate. The absence of a first place therefore does not mean an absence of support. It means that this support produced no local plurality in the file analysed.

The third concerns smaller parties. Their small provincial weight does not justify removing them from detailed tables. An honest analysis must keep their candidates and votes, notably to understand the full distribution in a close race.

2. Compared with 2022: the CAQ's reversal and gains for three parties

Sources: official 2022 archives and 2026 feed. Changes are calculated on vote counts, before percentages are rounded. [S01; S02]

This is a comparison between two general elections. It does not measure individual vote transfers: we cannot conclude that the votes lost by the CAQ are precisely those gained by any other party. Registered voters, turnout, new registrations and individual choices have also changed.

The electoral map also goes from 125 to 127 ridings. Provincial totals remain comparable as overall results, but a local comparison requires checking boundaries. An identical or similar name is not enough to establish that the same territory is being compared. 2022 results, results transposed onto another map and estimates must remain distinct. [S01; S02; S06]

Parti Québécois: strong growth, representation concentrated in several areas

The PQ receives almost twice as many votes as in 2022. Its share rises by 13.41 points. The picture goes from three seats won in 2022 to 59 ridings led in the 2026 feed.

Its strength shows in several territorial groupings: the 18 ridings of the Laurentides-Lanaudière grouping, the five of Saguenay–Lac-Saint-Jean and the six of the Bas-Saint-Laurent–Gaspésie–Îles-de-la-Madeleine grouping put it first. By contrast, it leads in none of the 11 ridings of the Capitale-Nationale grouping. [S01; S07; LQV calculations]

This contrast describes an electoral geography. It does not, on its own, explain the motivations behind that geography. A PQ win is not a measure of support for each of its proposals, notably a referendum: a ballot is cast for a candidate and a set of choices, not on a single referendum question.

Liberal Party: more votes and several centres of representation

The PLQ gains 445,974 votes compared with 2022 and rises by 9.84 points. It leads in 19 of the 27 ridings of the Île-de-Montréal grouping, in all six in Laval and in four of the five in the Outaouais. It also wins eight first places in Montérégie. [S01; S02; S07]

These results show that its representation is not limited to a single territorial grouping. They do not, however, establish the language profile, age or income of its voters. Describing the characteristics of a territory and attributing a choice to each resident are two different exercises.

Its margins also vary widely. In Vachon, its lead is 14 votes. In Saint-Henri–Sainte-Anne, it is 39. The provincial total of 40 first places therefore brings together results that are much closer in some ridings than in others.

Conservative Party: a significant entry into the representation picture

The PCQ wins 354,378 more votes than in 2022. Its share rises by 7.76 points. It leads in 19 ridings, after winning no seats in the previous general election. [S01; S02]

Fifteen of its 19 first places are in the Capitale-Nationale and Chaudière-Appalaches groupings: eight in the former, seven in the latter. In Chaudière-Appalaches, it wins 51.67% of aggregated valid votes. In Capitale-Nationale, it wins 36.23%, with eight first places out of 11. [S01; S07; LQV calculations]

This is a measurable territorial concentration. The data do not, however, show that its voters all chose the same political priority. The rise in support and its conversion into first places are established; a causal explanation requires other sources.

Coalition Avenir Québec: a drop in support amplified in representation

The CAQ loses 1,114,053 votes compared with 2022. Its provincial share falls by 27.64 points. The 90 seats won in the previous election contrast with no riding led in the current feed. [S01; S02]

The result is significant on both levels: lower votes and no first place. But the two measures are not interchangeable. The CAQ still has 571,520 votes and finishes second in 20 ridings. The analysis must keep these facts, even when attention turns to the size of its decline. [S01; LQV calculations]

The file makes it possible to observe the decline. It does not make it possible to attribute the whole result to a single person, a debate or a specific controversy. Such an explanation would require a documented timeline, opinion measures and a clear distinction between correlation and causation.

Québec solidaire: nine first places despite a lower vote

QS receives 102,359 fewer votes than in 2022 and loses about three points of provincial share. It leads in nine ridings, compared with 11 seats won in the previous general election. [S01; S02]

Its first places are spread across six ridings on the island of Montreal, two in Capitale-Nationale and one in Estrie–Centre-du-Québec. Margins and percentages vary considerably. Ruba Ghazal wins 59.20% in Mercier; Stéphanie Vachon leads in Sherbrooke with 27.16% and a 481-vote lead. [S01; S07]

The picture shows both a provincial decline and the persistence of local strongholds. Reducing this to a single verdict of “success” or “failure” would lose one of those dimensions.

3. Why votes and representation tell two stories

In each riding, the ranking depends on local votes. The provincial count then adds up those votes, but does not redistribute first places in proportion to their total.

Calculation: share of ridings led minus share of valid votes. This table measures a gap; it is not a simulation of another electoral system. [S01; LQV calculations]

A specific case illustrates local plurality: in Vachon, Affine Lwalalika comes first with 27.53% of the vote, ahead of Stephan Fogaing. The 14-vote lead is enough to separate the two candidates in the file's ranking. The other votes are spread among the other candidates.

Of the 127 ridings, 100 put a candidate with less than 50% of valid votes in the lead. In eight cases, that candidate is even below 30%. This does not make those ballots any less valid; it describes fragmented competition among several candidates. [S01; LQV calculations]

It would be misleading to add up the votes of several parties and claim to know who would have won under another voting system. A different system can change the candidates, strategies, alliances and voters' choices. A simple mathematical redistribution does not capture those effects.

4. The geography of the result

The groupings below are the territorial groupings used by Le Québec Vote. Some combine several administrative regions; they must not be presented as the 17 official administrative regions. Each riding is counted only once. [S07]

The CAQ is at zero in each of these groupings. Data and totals checked against the official feed. [S01; S07; LQV calculations]

Uniform first places across a territory do not mean voters were unanimous. In Laurentides-Lanaudière, the PQ comes first in all 18 ridings with 39.07% of aggregated valid votes. In Laval, the PLQ is first in all six with 41.24%. In Chaudière-Appalaches, the PCQ is first in all seven with 51.67%. Three uniform tables of first places thus correspond to very different levels of support. [S01; LQV calculations]

Estrie–Centre-du-Québec stands out because all four parties lead in at least one riding. Montérégie has 12 PQ first places and eight Liberal ones. These distributions call for a local reading: the aggregate must not erase close races or differences between municipalities.

5. Close races to watch

A close race calls for attention to corrections and to any official procedures. It does not prove that a recount was requested or that an irregularity exists.

Sorted by increasing absolute gap. Points are calculated from votes, then rounded. Names are kept as they appear in the feed. [S01; LQV calculations]

Papineau and La Prairie complete the list of ten races under two points, with gaps of 554 and 651 votes respectively. This shows why it is useful to display both measures: a gap in votes does not carry the same relative weight in a riding with few or many valid votes.

The official addition of votes checks the statements transmitted; a judicial recount recounts the ballots under a separate procedure. Simply counting every polling station is therefore not the same as completing every step. [S04]

6. Results of the five leaders followed during the campaign

Ridings and figures come from the 2026 feed, without automatically carrying over a 2022 candidacy. The order follows the provincial vote table. [S01]

The Trois-Rivières case links a personal result to a close race: Christine Fréchette finishes behind Jean Pellerin by 499 votes. Each leader's local result should nonetheless be distinguished from their party's provincial result.

7. Turnout: a modest provincial increase, wide local differences

The feed gives a provincial rate of 66.89%, compared with 66.15% in the 2022 archives. The difference is about 0.75 points. It also reports 4,345,857 votes cast out of 6,495,984 registered voters. [S01; S02]

Methodological note: the rate provided in 2026 and the ratio of the two totals do not match exactly. The calculated ratio gives 66.90%, about 0.0073 points higher. We keep the figure provided by the source and document this small difference; we infer no irregularity from it. The feed's documentation describes the rate as an estimate during the count. [S01; S03; LQV calculations]

Ranking limited to the rates reported in the snapshot of the 127 ridings. [S01]

These differences matter, but they do not demonstrate their cause. Access to voting, mobility, demographic composition, local competition or registration could be studied with additional data. None of these hypotheses becomes an established explanation merely because a rate is lower.

The 63,306 rejected ballots represent about 1.46% of votes cast, compared with about 1.35% in 2022. A rejected ballot should not automatically be described as a protest vote: the file does not provide the intention behind its rejection. [S01; S02; LQV calculations]

8. What the balance of power allows us to say about what comes next

With 127 ridings, the arithmetic threshold for an absolute majority is 64. The PQ's 59 first places are five short of that threshold. This describes the picture produced by the vote; it is not an announcement of a swearing-in or a confirmation of how the future government will be organized. [S01; LQV calculations]

In a minority government, parliamentary support becomes decisive. The government must retain the confidence of the Assembly. A defeat on an ordinary bill does not automatically amount to a loss of confidence; certain motions expressly engage that responsibility. Abstentions, members' attendance and voting rules also matter. All future votes therefore cannot be reduced to “five votes short.” [S05]

Tables adding up parties' totals prove neither a coalition, nor an agreement, nor a common program. Announcements, tabled texts and actual votes will have to be checked. Monitoring Parliament must start from that evidence.

9. After the promises: the evidence to look for

Le Québec Vote already has a promise tracker and documented issues. Their usefulness continues after the vote: comparing a campaign commitment with a bill, a budget or a measure that has actually taken effect.

The PQ's October 3 release announces priorities for the start of a possible mandate, notably on homelessness, taxation, government management, immigration and education. It is a source of promises, not proof that they have been carried out. [S08]

For each commitment, the follow-up questions are concrete: which text or decision implements it? What funding is authorized? When does it take effect? Who is eligible? What results can be measured?

An announcement, a tabled bill, an adopted law, voted appropriations and a service actually available are distinct stages. Dates relating to the mandate will have to be calculated from the documented starting point, without arbitrarily turning the day after the vote into “day 1.”

The promises of parties that do not govern should remain accessible. They can be used to follow their proposals and their votes. Their non-fulfilment by the government should not be attributed, as a failure in the exercise of power, to a party that does not exercise it.

10. What this analysis cannot conclude

The vote does not separately measure support for a tax cut, a health policy or independence. Local results do not reveal how 18-to-25-year-olds, parents or retirees voted. A young territory or one with many renters does not allow the same choice to be attributed to everyone who lives there.

Nor can we determine from these figures alone which debate changed the result, quantify an individual strategic vote or describe a close race as suspicious. These questions require separate evidence.

Our Generations section lets readers explore commitments according to life situations. It must remain a tool for reading promises, without becoming an invented portrait of voting by age.

11. More than half of the votes went to other candidates

Adding up the votes of the 127 candidates who finished first gives 1,786,239 votes. The other candidates together received 2,496,312 votes: 58.29% of all valid ballots. This figure describes how votes are distributed between first places and other ranks in each riding. It does not mean that 58.29% of people voted against a future government: the local first-placed candidate can belong to any party. [S01; LQV calculations]

Nor does it mean that these votes are useless. They count in the result and in measuring party support. But in the table of first places, they do not give first position to the candidate chosen. This helps explain why the result in ridings and the result in votes describe two different dimensions of the same vote.

Each row uses its own party's votes as the denominator; the provincial 58.29% also includes the other parties and independents. The table measures neither voter satisfaction nor the number of people who consider themselves represented. [S01; LQV calculations]

The contrast between the PCQ and QS is instructive: their provincial shares are not enough to explain their number of first places. We also need to look at where their support lies and which other parties finish ahead of them locally. The same number of votes can be spread across many ridings or concentrated in a few; the provincial total does not contain that information.

12. The CAQ: no first place, but different local situations

The zero in the provincial table sums up the CAQ's first places. It does not, by itself, describe the position of each of its 127 candidates. The local ranking gives a more precise picture: 20 second places, 33 third, 44 fourth and 30 fifth. The CAQ therefore finishes fourth or fifth in 74 ridings, or 58.27% of its candidates. [S01; LQV calculations]

Every party in this table has a candidate in each of the 127 ridings. This is the rank in the 2026 vote, not a ranking of their ability to win the next election. A second place with a large gap to close is not the same as a second place a few dozen votes behind. [S01; LQV calculations]

The PQ combines 59 first places and 41 seconds: it is among the top two candidates in 100 ridings. For the PLQ, that total is 61; for the PCQ, 51; for QS, 22; for the CAQ, 20. This shows how widely they are present in the top local positions, without predicting any future shift of voters.

This reading also avoids turning the CAQ's provincial decline into a single explanation. The loss of votes is measurable. The respective weight of the campaign, the leaders, the government's record or the issues requires separate data. The result does not allow this loss to be divided among those causes.

13. Where are the votes, and where are the first places?

To understand a party's geography, two questions must be separated: what share of a territory's votes does it win? And what share of its own provincial votes comes from that territory? The first percentage describes local competition; the second describes the distribution of its support. Population and the number of ridings differ between groupings, which prevents comparing their volumes as if they were the same size.

The PCQ receives 162,732 votes in the Capitale-Nationale grouping and 143,795 in Chaudière-Appalaches. Together, these 306,527 votes represent 34.63% of its provincial votes. Yet these two groupings account for 15 of its 19 first places, or 78.95%. The other groupings therefore account for about two-thirds of its votes and four first places. This is a concrete illustration of how concentrated its first positions are. [S01; LQV calculations]

The PQ receives 254,595 votes in Laurentides-Lanaudière and 240,060 in Montérégie: together, 41.22% of its provincial votes. It comes first there in 30 ridings out of its 59 first places. The PLQ receives 329,417 votes on the island of Montreal, or 31.76% of its provincial votes, and wins 19 of its 40 first places there. [S01; LQV calculations]

QS receives 194,865 votes on the island of Montreal, or 36.62% of its provincial votes, and has six of its nine first places there. A majority of its votes therefore come from outside that grouping, where it comes first in three ridings. Describing QS only through its Montreal first places would erase that part of its support. [S01; LQV calculations]

For the CAQ, Montérégie accounts for 123,018 votes and Laurentides-Lanaudière for 103,759. Together, these two groupings provide 39.68% of its provincial votes, without a first place. Their weight in its total indicates where its 2026 support is located; it demonstrates neither a transfer to another party nor the cause of its decline.

These groupings are the ones used in the site's directory, including composite groupings. They do not correspond to an analysis of the 17 administrative regions. A map for visitors will have to name them as such and allow each riding to be opened. The provincial table, the territorial view and the local profile are complementary: each uses a different denominator.

14. The last two verified polls and the result

The voting intentions published by Pallas on October 4 and by Léger on October 3 placed the five main parties in the same order as the provincial result in votes. This observation concerns these two verified waves, not all of the campaign's polls. [S10; S11; S12]

Sources and bases: Pallas, 1,137 respondents, of whom 1,114 were decided or leaning toward a choice; Léger, 1,000 respondents, decided base of 892; results, 4,282,551 valid votes. Polls publish rounded numbers. [S01; S10; S12]

For these five parties, the observed differences are all under two points. QS finishes 1.43 points above its published intentions in both waves; the CAQ finishes 1.65 points below Pallas and 0.65 points below Léger. These are descriptive gaps between a pre-election measure and the result. They prove neither a last-minute shift nor strategic voting. [S01; S10; S12; LQV calculations]

Pallas used automated calls to landlines and mobile phones. Léger used its LEO online panel, which is non-probabilistic: the stated comparative margin of 3.1 points is not a conventional margin for the panel. Methods, fieldwork and bases differ. A single wave per firm is not enough to rank their reliability.

Provincial closeness does not guarantee local accuracy: this table validates no riding projection. It cannot tell us the outcome of a close race or the actual choices of an age group. The comparison is useful for placing the vote in the context of end-of-campaign expectations, within these limits.

And in your riding?

The provincial result gives the balance of power. The local result shows who leads where you live and by how much. The detailed results, the riding profiles, the promise tracker and the issue comparison tool let you keep reading.

Questions about housing, care or transportation in your community remain relevant after the election. Following them will mean connecting documented commitments to public decisions, while keeping sources and corrections.

Sources and method

References S01 to S12 point to the numbered source list below, with their links, dates and limits. The calculations cover the 127 ridings and 908 candidates in the 2026 feed. Names, votes and first places were checked using their official identifiers. Totals by candidate, riding and party match the provincial totals. The files analysed and the calculations are archived and reproducible; on October 6, 2026, the archived file was compared with the online official feed and found identical.

Comprend des données ouvertes octroyées sous la licence d'utilisation des données ouvertes du directeur général des élections disponible à l'adresse Web dgeq.org. L'octroi de la licence n'implique aucune approbation par le directeur général des élections de l'utilisation des données ouvertes qui en est faite.

(Includes open data granted under the Chief Electoral Officer's open data licence, available at dgeq.org. The granting of the licence does not imply any endorsement by the Chief Electoral Officer of the use made of the open data. Unofficial English rendering of the notice above.)

Le Québec Vote is independent of Élections Québec. Last verified: October 6, 2026. The text presumes no recount, outcome of a legal challenge or change of government.