Reports on the relationship between leptin and physical activity (PA) at the population level are scarce. The present study examined the relationship between serum leptin concentrations and PA in a nationally representative sample of 3001 Iranian adults aged 25 to 64 years. Data of our third national surveillance of risk factors of noncommunicable diseases were analyzed. Using the Global Physical Activity Questionnaire, the duration and intensity of PA were evaluated in 3 domains: work, commuting, and recreation. Total PA was calculated using metabolic equivalents for PA intensity. Serum leptin was measured with an enzyme-linked immunosorbent assay. After adjustment for age, area of residence, smoking, body mass index, and waist circumference, total PA (r = -0.129, P = .038 in men and r = -0.226, P = .006 in women), the duration of vigorous-intensity activity (r = -0.120, P = .044 in men and r = -0.154, P = .019 in women), the duration of moderate-intensity activity (r = -0.114, P = .047 in men and r = -0.160, P = .018 in women), and time spent on sedentary behaviors (r = 0.194, P = .014 in men and r = -0.204, P = .007 in women) were significantly correlated with serum leptin. In both sexes, participants in higher categories of PA had significantly lower serum leptin levels. In conclusion, our results demonstrated an inverse association between leptin concentrations and PA independent of age, sex, smoking, and body adiposity. Our results point to the regulatory effects of PA on serum leptin.