We conducted a multicity time-series study using monitoring data to assess seasonal patterns of short-term ozone-mortality association among elderly aged 65 years and over in Japan. Daily exposure to ambient ozone was computed using hourly measurements of photochemical oxidants available at multiple monitoring stations in each city. Effects of ozone on daily all-cause non-accidental, cardiovascular, and respiratory mortality were estimated using distributed lag linear models, controlling for confounding by temporal, day of the week, temperature, and flu epidemics. City-level effect estimates were combined using inverse variance meta-analysis. In spring and autumn, a 10-ppbv increase of daily maximum 8-h average ozone concentration in the previous 3 days was associated with 0.69 % (95 % confidence interval (CI): 0.27-1.10), 1.07 % (0.34-1.82), and 1.77 % (0.78-2.77) increases in daily all-cause, cardiovascular, and respiratory mortality, respectively. Forward displacement of respiratory mortality was large during the cold season despite lower ozone concentration. Results were generally independent of fine particulate matter and nitrogen dioxide. Findings suggest significant mortality effects of short-term ozone exposure among the elderly during the moderate season. Those with underlying respiratory diseases were susceptible, even during winter.