This study aimed to investigate whether social isolation is associated with mortality, together with the effect of the Great East Japan Earthquake on mortality, due to the social isolation of community residents living in the affected areas, using data from the Tohoku Medical Megabank Project Community-Based Cohort Study. A total of 22,933 participants (8059 men and 14,874 women), who were free from cancer and cardiovascular disease, were followed up with death as an endpoint for five years. Social isolation was assessed using the Lubben Social Network Scale (cut-off, 11/12). Using Cox proportional hazards models, hazard ratios (HRs) of total mortality and 95% confidence intervals (CIs) associated with social isolation (no isolation as the reference) were estimated. The latter was significantly associated with an increased risk of total mortality (1.38 (1.04–1.83) in men and 1.49 (1.02–2.19) in women). Moreover, among those with social isolation, the risk of mortality was significantly higher, especially for women with house damage and men who had experienced a death in the family. The disaster may have raised the risk of mortality due to social isolation.