Queensland, February 2026

01

One party moved and three did not

A change between the October and February waves carries about ±5.2 points at 95 per cent, from effective samples of 747 and 702. One Nation's rise of seven is the only movement that clears it.

Each bar is the change in a party's first preference from October 2025 to February 2026, in points. The pale band is the ±5.2 points a change of that size carries.

The four falls sum to seven, which is One Nation's gain, so nothing entered or left the poll between the two waves. The LNP supplied three of those points, the Greens two, and Labor and the unnamed others one each. Only the LNP's loss is large enough to be worth a second look on its own, and even that sits inside the interval: a three-point fall against ±5.2 is consistent with no change at all.

What the wave does establish is One Nation's level. Its 21 per cent carries an interval of about ±3.0 on this sample, so the party is somewhere between 18 and 24, and that is a different order of support from the 8.0 it took at the 2024 election. The rise since the election is 13.0 points and it has happened across two waves rather than one, which is a steadier picture than a single jump would be.

02

The LNP lost three points of primary vote and gained two of two-party preferred

DemosAU publishes the flow rates it applies. Running them over the February primaries reproduces the published figure and shows where the gain comes from.

The flow rates are DemosAU's, printed in its methodology statement: Greens 81 per cent to Labor, One Nation 74 per cent to the LNP, others 56 per cent to the LNP. Queensland's full preferential system distributes every ballot, so the two columns sum to 100 and nothing exhausts. Each cell is rounded to one decimal place and the totals are struck on unrounded values. The reproduction gives 55.4 against a published 56, the difference being what rounding the primary votes to whole numbers costs.

One Nation's 21 points deliver 15.5 to the LNP and 5.5 to Labor, a net of ten points in the LNP's favour from a party it is losing primary votes to. In October the same calculation on a One Nation vote of 14 delivered a net of about seven. The movement between the two waves decomposes cleanly: One Nation's rise of seven adds 5.18 points to the LNP's two-party share, the LNP's own fall of three subtracts 3, the Greens' fall of two subtracts 0.38 and the others' fall of one subtracts 0.56, for a net of 1.24. The published figures move from 54 to 56 and the same flows reproduce them as 54.1 and 55.4.

A government in this position is being told two things at once. It is less popular than it was, by a margin too small for one wave to confirm, and it is better placed to win the next election than it was, by a margin that is real under the current flow assumption. Both are true, and the second depends entirely on One Nation preferences continuing to behave as they did in 2024. A reader who takes 56–44 as evidence the government is recovering has read the wrong line; a reader who takes the primary vote as evidence it is in trouble has read the wrong line too.

03

One Nation outpolls the LNP among voters aged 35 to 54

DemosAU publishes no cohort sample sizes for this wave. If the three-way splits each hold about a third of the effective sample of 702, a gap between two of them carries roughly ±8.3 points.

Every row sums to 100 in the report. These figures are read from the demographic chart on the report's own page 9. Cohort figures rest on subsamples and carry wider margins than the ±3.7% statewide topline.

One Nation's 30 per cent among voters aged 35 to 54 is the first cohort in this series where it leads the party of government, by one point over the LNP's 29. The one point establishes nothing, but the level does: a minor party taking three in ten of the largest working-age cohort is a different proposition from one taking a fifth of the state. In regional Queensland it takes 29 against the LNP's 33, and among renters 25 against 22.

The LNP's own vote has become sharply conditional on age and assets. It takes 47 per cent of the over-55s and 22 of the under-35s, 46 of outright owners and 22 of renters, and 50 per cent of those earning over $125,000 against 33 in both lower income bands. Each of those gaps clears the roughly ±8.3 points a cohort comparison carries here. Brisbane remains the exception it was in October, with the Greens on 15 and One Nation on 12, the only cohort where that ordering holds.

04

February against October and the 2024 election

Percentage points, first preferences. Sixteen months after the election the government's primary vote is 7.5 points lower and its two-party share is 2.2 points higher.

The February column is as the June report publishes it in its own comparison charts; the February report carries its topline pages as images with no text layer. The 2024 and October columns are from the October report and the June report's election baselines, which agree.

Read down the One Nation row and the term has one story: 8.0, then 14, then 21. Read the two-party row and it has another: 53.8, then 54, then 56. Both rows are the same voters. The party of government has lost nearly a fifth of its primary vote since the election and improved its position on the measure that decides seats, and the June wave carries both further in the same directions.