BACKGROUND AND AIM[|]In previous analyses, we found that surrounding greenness was associated with better mental health in urban and peri-urban areas. Physical activity has been suggested to be a mediator of this association, but existing literature remains inconsistent. We aimed to assess the causal mediating effect of physical activity on this association in a large French population-based cohort.[¤]METHOD[|]77,000 participants from the French CONSTANCES cohort (inclusion 2012-2018), with available data from the 2019 follow-up questionnaires and living in peri-urban (N=46,289) and urban (N= 30,711) areas were included. Physical activity was measured at inclusion with a total score ranging from 0 to 6. Mental health was measured using the General Health questionnaire (GHQ-12) continuous score. Surrounding greenness was quantified by the normalized difference vegetation index (NDVI) within 100m and 300m buffers around the residential address of each participant in 2019. The mediating effect of physical activity was evaluated using adjusted causal mediation models for each degree of urbanity and for each buffer.[¤]RESULTS[|]54% were women and mean age was around 49. An increase of NDVI was associated with a better mental health in urban areas and peri-urban areas. No mediating effect of physical-activity on the association between surrounding greenness within a 300m-NDVI buffer and mental health was found in peri-urban areas (indirect effect β= -0.01 (-0.03,0.01), p=0.53; total effect β=-0.34 (-0.63, -0.06), p=0.02) nor in urban areas (indirect effect β=0.02 (-0.00,0.04), p=0.11; total effect β= -0.42(-0.73, -0.14), p=0.01). Similar results were found with greenness within 100m-NDVI buffer (all p for indirect effect > 0.05).[¤]CONCLUSIONS[|]Physical activity does not mediate the association between greenness and mental health in peri-urban or urban areas. Further evaluation of other mediators or greenness metrics is needed to understand associations of greenness with mental health.