Background Depression is ranked as the leading cause of global disability. Physical activity has been identified as a relevant strategy to reduce the incidence of depression. However, specific aspects of physical activity (eg, type of activity or combination) in decreasing the risk of depression have been underexplored. We aimed to investigate the associations of the type of activity, the number of activities participated in, and the combination of different activity types with the risk of depression onset. Methods We included 161 023 participants aged 38–70 years from the UK Biobank study. At baseline, participants reported their participation in walking, light do-it-yourself (DIY) activities, heavy DIY activities, sports, and other exercises. Depression onset was extracted from primary care and hospital admission records. Cox proportional models were performed, with a 2-year landmark analysis. Analyses were controlled by sociodemographic, lifestyle, and health-related covariables, as well as by total time in physical activity and frequency of physical activity. Findings A total of 88 723 (55∙1%) of 161 023 participants were women and 72 300 (44∙9%) were men. 5401 (3∙4%) of 161 023 participants developed depression over a median follow-up of 10∙01 years (IQR 9∙30–10∙83). Compared with no physical activity, the hazard ratio (HR) for depression onset was 0∙79 (95% CI 0∙71–0∙88) for participation in walking, 0∙81 (0∙72–0∙91) for light DIY, 0∙72 (0∙63–0∙82) for heavy DIY, 0∙62 (0∙49–0∙78) for sports, and 0∙73 (0∙65–0∙82) for others exercises. The risk of depression decreased by 19% (95% CI 10–27) in those who engaged in one activity, 25% (16–33) in two activities, 31% (21–39) in three activities, 41% (31–49) in four activities, and 58% (44–69) in five activities. Among the ten most frequent combinations, walking for pleasure, light DIY, heavy DIY, and other exercises showed the lowest risk for depression onset (HR 0∙57 [95% CI 0∙48–0∙68]). Interpretation Participation in low-intensity activities, such as walking or light DIY, was associated with a lower risk of depression in comparison with no physical activity participation. Encouraging people to combine different physical activities could lower the risk of depression further.