Background: Respiratory diseases are directly affected by reduced exposure to green areas and air pollutants, both adversely affecting health. Aims: to evaluate whether self-reported asthma and COPD diagnoses are associated with green areas and PM2.5. Methods: 3,396 participants of the ELSA-BRASIL cohort study from São Paulo City with complete data were enrolled during 2008-2010. Asthma (n=374) and COPD (n=74) were defined by self-report at baseline. 2010 orthophoto was used to classify the land cover, and classification was performed using the random forest algorithm method. The images were classified as tree canopy, grass, bare ground, cement floor, pool, shade, white roof, gray roof, ceramic roof, dark roof, asphalt, and river/lake. Global PM2.5 concentrations were estimated using information from satellites applying a Geographically Weighted Regression(GWR). We calculated each participant's annual mean values and built binary logistic regression models to verify whether asthma or COPD diagnosis at baseline was associated with green area variables and PM2.5. Results: Participants had similar demographic characteristics: 60% women, age 52±9 years, and 60% white race. There were no differences in the amount of green and land cover for almost all parameters except for the white roof in crude and adjusted models. When we restricted analyses to participants living in the same place for five or more years (n=2,814), the association between white roofs and asthma vanished. Conclusion: Asthma and COPD were not associated with the density of green areas and indirect assessment of atmospheric pollution of PM2.5 in São Paulo City. Sponsorship: FAPESP(2018/02537-5 and 2019/26449-0), LIM20-HC-FMUSP, CNPq(01060115.00SP)