Datafolha indicates Lula with 38% and Flávio Bolsonaro with 33% in the first round; Augusto Cury rises to 8%

New research by Datafolha shows that President Lula (PT) has 38% of voting intentions in the first round, compared to 33% for Flávio Bolsonaro (PL). It also confirms the rise of Augusto Cury (Avante), who rose six points and reached 8%.

As a result, the difference between the incumbent and the main challenger reached five points, half of what was recorded in May.

Then come Ronaldo Caiado (PSD), with 4%, Renan Santos (Missão), with 3%, and Romeu Zema (Novo), with 2%. Samara (UP), Edmilson Costa (PCB) and Rui Costa Pimenta (PCO) have 1%. The rest did not score. 6% declare a blank or null vote, and 3% say they are undecided.

In the previous survey, two weeks ago, the PT member had 39% and the PL senator 33% in the first round simulation. Caiado scored 5%, Renan, 4%, Zema, 3%, and Cury, 2%.

In the second round, the president numerically surpasses his predecessor Jair Bolsonaro's son by 46% to 44%, a technical tie due to the survey's margin of error, plus or minus two points. He beats Caiado by 46% to 41% (was 47% to 40%), Zema by 48% to 39% (was 48% to 38%) and Renan by 47% to 38% (was 47% to 37%).

In the second round scenario two weeks ago, Lula had 47%, compared to Flávio's 43%. In all other scenarios tested, there was a decrease within the margin of error of the PT member's advantage. The rejection of both continues to be heavy: 47% say they will not vote at all for the senator and 45% for the president.

Datafolha interviewed 2,002 voters in 125 cities on Tuesday (1st) and Wednesday (2nd). The research, contracted by TV Globo and Folha, is registered with the Superior Electoral Court with code BR-03669/2026.

These data are from the scenario researched by Datafolha without Pablo Marçal (PRTB), who is ineligible, but managed to register his candidacy for the Presidency — which tends to be blocked when judged by the TSE.

When it is tested, the race does not change. Marçal scores the same 2% as in the previous survey, with minimal changes to the rest of the table.

In the spontaneous research of the main scenario, without the influencer, the scenario is one of stability at the top: Lula has 31% of citations without the interviewee seeing the list of candidates, and Flávio scores 22%. Cury goes from 1% to 5%, and the number of undecided people, important given how close the election is, drops from 32% to 26%.

From the point of view of the electorate profile, the profile of those who vote for the leaders is stable.

An exception is a reversal within the margin of error in favor of Lula in the Southeast, where the president now scores 36% to 34% compared to Flávio, who recorded the same score for himself two weeks ago. The region concentrates 42% of the electorate in the Datafolha sample.

Furthermore, Lula remains ahead of Flávio, in the main groups, among the 50% poorest (47% to 27%), among the 30% least educated (51% to 25%), among the 27% of Northeasterners (53% to 22%) and among the 49% of Catholics (44% to 29%).

The senator surpasses the president among the 25% with higher education (37% to 30%), the 33% of middle class people who earn 2 to 5 minimum wages (38% to 32%), the 15% of southerners (44% to 27%), the 16% of residents of the North/Mid-West (39% to 33%) and the 27% of evangelicals (45% to 24%).

After months of sluggishness, with the dispute appearing to be limited to the top two, the presidential race is experiencing a moment of turmoil.

The current survey is the first by Datafolha after the start of free electoral hours. During the period there was also the empty inaugural debate of the campaign on TV Band and presidential hearings on TV Globo.

Furthermore, researchers took to the streets amid the impact of the crisis on the STF (Supreme Federal Court). It is important to note that the survey was already underway when the scandal broke, so its impact cannot be read as integral.

The revelation of compromising messages between the former owner of Banco Master, Daniel Vorcaro, and minister Alexandre de Moraes will become propaganda ammunition for both Lula and Flávio and tends to be unfavorable for both in the end.

The PT member will be accused of the tacit alliance with Moraes, Jair Bolsonaro's legal tormentor, while the Rio de Janeiro senator will have explored his connection with Vorcaro, from whom he asked for money to allegedly finance a film about his father.

Before the crisis broke out, other surveys had indicated the rise of outsider Cury, who has 35 seconds on TV and radio. The presumed reason was the flood of quotes on the networks caused by influencers in favor of his name.

In two weeks from August 16, the writer saw his Instagram account gain another 2.5 million followers, something unusual. According to him, the movement is spontaneous. His rivals suspect some boosting mechanism with robots or purchase of support, which Cury denies.

In any case, he achieved an unexpected highlight and suggested some dynamics to the fierce election in the first round. Datafolha had already recorded this research when the movement became evident, and did not simulate the dispute in the second round between Lula and Cury.

From the point of view of the electorate profile, Cury is better among those aged 25 to 34 (14%), those with a higher degree (14%) and those who say they are not aligned with Bolsonarism or PTism (18%). Its rejection is quite low: 9%.