BACKGROUND: The impact of length of hospital stay on activities of daily living (ADLs) has not specifically been investigated among dialysis patients. Therefore, we attempt to verify the association between the length of hospital stay and the decline in ADLs among hemodialysis patients. METHODS: This prospective cohort study used data from the Japanese Dialysis Outcomes and Practice Patterns Study (J-DOPPS). We included 2442 hemodialysis patients aged ≥40 years from the J-DOPPS phase V (2012-2015) and subsequently excluded those who had already lost basic activities of daily living (BADLs) as demonstrated by dependency in at least three of the five BADLs at baseline and for whom changes in ADLs had been evaluated for less than 90 days. The main exposure was the cumulative length of hospital stay during the follow-up period. The primary outcomes were a decline in at least one of the five BADLs and eight instrumental activities of daily living (IADLs). We compared risk ratios (RRs) for 30-day increments for hospital stays with 10-year increments for age and having diabetes. RESULTS: A total of 849 patients were included in the statistical analysis. The cumulative length of hospital stay was significantly associated with a risk of decline in ADLs (adjusted RRs [95% confidence intervals] per 30-day increments: 1.42 [1.15 to 1.75] for BADLs, 1.38 [1.13 to 1.68] for IADLs). The adjusted RRs [95% CI] for 10-year increments in age were 1.20 [0.96 to 1.50] and 1.21 [1.00 to 1.47]. The adjusted RRs [95% CI] for having diabetes were 1.36 [0.97 to 1.91] for BADLs and 1.38 [1.04 to 1.84] for IADLs. CONCLUSION: The impact of a 30-day increment in the cumulative length of hospital stay on the decline in ADLs was comparable to that of a 10-year increase in age and having diabetes.