Summary
An AtlasIntel/Bloomberg poll released on Wednesday indicates President Luiz Inacio Lula da Silva leading opposition Senator Flavio Bolsonaro in a potential second-round vote ahead of October’s presidential election. The survey places Lula at 49.2% and Bolsonaro at 42.9% in a simulated runoff, and also provides first-round vote shares across several named candidates.
Second-round snapshot
In the poll’s head-to-head scenario for a second-round ballot, Lula receives 49.2% of the projected vote, while Flavio Bolsonaro is shown with 42.9%. The gap is marginally larger than in June, when Lula registered 48.8% against Bolsonaro’s 42.3% in the same simulated matchup.
Notably, the proportion of respondents who were undecided or said they intended to cast blank or spoiled ballots declined to 7.9% in the latest survey, down from 8.9% in June.
First-round preferences
The poll also reports first-round support numbers. Lula tops that list with 44.9% of respondents indicating support, followed by Senator Bolsonaro at 35.8%. Other named candidates appear with smaller shares: Renan Santos at 7.8%, Ronaldo Caiado at 3.1%, and Romeu Zema at 2.8%.
Methodology
The survey was carried out between July 22 and July 27 and included 5,021 participants. The stated margin of error for the results is one percentage point in either direction.
Key points
- The AtlasIntel/Bloomberg poll shows Lula leading Bolsonaro in a second-round scenario, 49.2% to 42.9%.
- In first-round figures, Lula is reported at 44.9% and Bolsonaro at 35.8%; other candidates register single-digit support.
- The sample size was 5,021 respondents, with a margin of error of plus or minus one percentage point; undecided or blank/invalid voting intentions fell to 7.9% from 8.9% in June.
Risks and uncertainties
- The margin of error of one percentage point means actual support could vary slightly above or below the reported figures.
- The remaining 7.9% of respondents who were undecided or planned blank/spoiled ballots represent a portion of the electorate that could influence outcomes if preferences consolidate differently.
- As with any snapshot poll, the results reflect views captured between July 22 and July 27 and may change before the October vote.
This report presents the poll results and methodology without inferring causes or projecting outcomes beyond what the survey data show.