The president (PT) leads all first round scenarios for the 2026 presidential election and appears technically tied with the governor of , (Republicans), in a possible second round, according to a survey released this Tuesday (13) by the Meio portal in partnership with the research institute Ideia.
In the five stimulated first-round scenarios tested by the research, Lula maintains the lead. Against Tarcísio de Freitas, the PT member scores 40.2%, compared to 32.7% for the São Paulo governor. In one of the scenarios with (-RJ), the senator reaches 27.6% against 39.6% for the president.
In the dispute against (PL), the PT member has 40.1% against 29.7% for the former president’s wife (PL).
In a possible second round between Lula and Tarcísio, there is a technical tie within the margin of error: the president has 44.4%, while the governor of São Paulo has 42.1%. In the other second round scenarios tested, Lula appears ahead of his opponents. He scores 46% against 37%. (-PR), 46.3% against 36.5% of ( – GO), 46% against 39% of Michelle, 46.3% against 36.1% of (Novo – MG) and 46.2% against 36% of Flávio.
In the spontaneous question — in which interviewees answer “In 2026 we will have the presidency of Brazil, if the elections were held today who would you vote for?” — Lula appears with 32% of voting intentions. Former president Jair Bolsonaro, ineligible and sentenced to 27 years and three months in prison, has 9.5%, followed by Flávio Bolsonaro (6.6%), Tarcísio de Freitas (6.1%) and Michelle Bolsonaro (3.6%). The remaining names are below 2%.
Lula is also the most remembered candidate when participants were asked who they would not vote for at all. 40.8% indicated that they rejected the PT member. The president is followed by Flávio Bolsonaro (30%), Michelle (26.1%) and Tarcísio (16.2%).
Regarding the vote for president, 64.5% say they are decided, against 35.5% who say the opposite. Half of Brazilians think that Lula does not deserve to remain in power, 46.9% think the opposite and 3.1% do not know.
The survey interviewed 2,000 people between January 8 and 12, 2026, through telephone interviews. The margin of error is 2.2 percentage points, plus or minus, with a 95% confidence interval. The survey is registered with the Superior Electoral Court under protocol BR-06731/2026. Due to rounding, the sum of percentages may not add up to exactly 100%. The survey was paid for by Meio.
