Objective: The purpose of this study was to investigate the interaction between the G-protein beta-3 (GNB3) 825C4T polymorphism and physical activity in relation to prevalent obesity and hypertension. Research methods and procedures: The GNB3 825C4T genotype was measured in a sample of 14 716 African Americans (AAs) and whites from the Atherosclerosis Risk in Communities (ARIC) study, and logistic regression was used to test for genetic effects and gene-environment interactions. Results: The GNB3 825C4T variant was not independently associated with prevalent obesity or hypertension in either AA or whites. However, we observed a significant interaction (Po0.001) between this variant and physical activity in predicting obesity status in AAs. In AAs who were active, each 825T allele was associated with a 20% lower prevalence of obesity (odds ratio (OR) ¼ 0.80, 95% confidence interval (CI) ¼ 0.689-0.937, P ¼ 0.005), whereas each 825T allele was associated with a 23% greater prevalence of obesity for low-active individuals (OR ¼ 1.23, 95% CI ¼ 1.06-1.44, P ¼ 0.008). We also found a significant interaction between the GNB3 825C4T polymorphism, obesity status and physical activity in predicting hypertension in the AA subjects. AA homozygotes for the 825T allele who were both obese and had a low activity level were 2.7 times more likely to be hypertensive, compared to non-obese, active 825C homozygotes (OR ¼ 2.71, 95% CI ¼ 1.19-6.17, Po0.02). Discussion: Our findings suggest that the variation within the GNB3 gene may interact with physical activity level to influence obesity status and, together with obesity and physical activity, the GNB3 825C4T variant may influence hypertension prevalence in AAs.