Aim The National Health and Nutrition Survey, Japan, has annually monitored two indicators of physical activity in adults. They are contrasting in the association with age; the prevalence of exercise habit is lower and step counts are higher among younger participants. The present study aimed to examine the effects of rapid aging of the Japanese population and the lower participation rate among younger adults on the short‐term trend of two indicators of physical activity using tabulated data. Methods The prevalence of exercise habit and step counts by age groups (≥20 years) from 2003 to 2010 were estimated using tabulated data from the National Health and Nutrition Survey by calculating sex‐specific means weighted by age‐specific Japanese population data for each year (population‐weighted estimates) and for a fixed year (2005; age‐standardized estimates). Linear regression analyses were used to test the statistical significance of their trends. Results Statistically significant increasing trends in the prevalence of exercise habit were observed for the crude means (P = 0.029), the population‐weighted estimates (P = 0.007) and the age‐standardized estimates (P = 0.016) only in men. Statistically significant decreasing trends in the step counts were observed for the crude means (P = 0.006 in men and P = 0.033 in women) and the population‐weighted estimates (P = 0.008 in men and P = 0.049 in women) both in men and women, but for the age‐standardized estimates (P = 0.039) only in men. Conclusions The effects of rapid aging of the Japanese population and the lower participation rate among younger adults on the short‐term trend are not small, and age‐standardization is necessary to observe even the short‐term trend of physical activity data. Geriatr Gerontol Int 2017; ••: ••–••.