Importance Growing evidence suggests that physical activity protects against digestive system cancers (DSCs). However, it remains largely unknown whether consistently reaching the physical activity guidelines (≥7.5 metabolic equivalent task [MET]-hours/week) is associated with lower DSC risk or whether a much higher level is needed. Objective To examine the association between physical activity and DSC risk and mortality with a focus on the optimal amount and long-term consistency in performing the recommended level over a long period of time. Design, Setting, and Participants This population-based cohort study includes data from 3 large US prospective cohorts: Health Professionals Follow-Up Study, 1988-2020; Nurses’ Health Study, 1988-2021; and Nurses’ Health Study II, 1991-2021. Data analysis was performed between October 2024 and May 2025. Participants were men and women free from cancer and cardiovascular disease at baseline. Exposures Total levels of leisure-time physical activity were assessed with biennial validated questionnaires and expressed in MET-hours per week. Consistency was calculated as the percentage of follow-up years meeting the recommended level (≥7.5 MET-hours/week). Main Outcomes and Measures DSCs included cancers of digestive tract (mouth, throat, esophagus, stomach, small intestine, colon, and rectum) and digestive accessory organs (pancreas, gallbladder, and liver). Results During up to 32 years of follow-up of 231 067 men and women (median age, 43 [IQR, 36-55] years), a total of 6538 incident DSCs and 3791 DSC deaths were documented. Higher physical activity levels were associated with lower DSC risk (≥45 vs &lt;3 MET-hours/week; hazard ratio [HR], 0.83; 95% CI, 0.74-0.93; P &lt; .001 for trend) and mortality (HR, 0.72; 95% CI, 0.62-0.83; P &lt; .001 for trend). The inverse associations were evident for both digestive tract and accessory organ cancers: the HRs comparing greater than or equal to 45 vs 3 MET-hours/week were 0.85 (95% CI, 0.75-0.97) for digestive tract cancer risk and 0.73 (95% CI, 0.58-0.92) for digestive accessory organ cancer risk. The traditional dose-response analysis suggested the lowest DSC risk was achieved at approximately 50 MET-hours/week. However, when long-term consistency was considered, compared with those with minimal activity, consistently reaching the guideline at moderate levels (median, 16.9 [IQR, 13.6-20.5] MET-hours/week) over 3 decades was associated with substantial reductions in DSC risk (HR, 0.83; 95% CI, 0.75-0.90), whereas that performing much higher amounts (median, 38.5 [IQR, 28.5-53.8] MET-hours/week) was not associated with further benefit (HR, 0.87; 95% CI, 0.81-0.93). Conclusions and Relevance In this study, traditional dose-response analysis suggested that approximately 50 MET-hours/week was associated with optimal DSC risk reduction. After incorporating long-term consistency, maintenance of a moderate level approximately 17 MET-hours/week over 3 decades was associated with achieving optimal benefit in reducing DSC risk.