AIMS: We aimed to investigate the temporal risk patterns of severe hypovolemia induced by sodium-glucose cotransporter-2 inhibitors (SGLT2i) in patients with type 2 diabetes mellitus. MATERIALS AND METHODS: We conducted a self-controlled case series using claims data from Japan. Patients who were prescribed SGLT2i for treating type 2 diabetes mellitus and experienced severe hypovolemia were enrolled. The primary analysis evaluated the adjusted incidence rate ratios (IRRs) of exposure risk periods (Days 1 to 30, Days 31 to 90, Days 91 to 180 and Days ≥181) with their corresponding 95% confidence intervals (CIs), calculated using a multivariable conditional Poisson regression model, relative to that of the unexposed control period. RESULTS: A total of 1200 new users of SGLT2i with 1334 severe hypovolemia events were included. The median follow-up and treatment periods were 3.66 and 2.53 years, respectively. The cohort was predominantly male (78.4%) with a median age of 54.1 years. A higher risk of severe hypovolemia associated with SGLT2i was observed particularly in the first 30 days (adjusted IRR 7.39, 95% CI 6.09-8.96) of treatment initiation. Secondary analyses highlighted the first 22 to 28 days (adjusted IRR 15.24, 95% CI 11.92-19.48) of treatment as the highest risk period for severe hypovolemia. CONCLUSIONS: SGLT2i use in patients with type 2 diabetes mellitus was associated with a higher risk of severe hypovolemia, particularly within the first 30 days of treatment initiation, with the highest risk observed during 22 to 28 days.