The aim of the study was to determine any association of physical activity and Taq 1B polymorphism in the cholesteryl ester transfer protein gene on high-density lipoprotein (HDL) cholesterol. Five hundred and four subjects, 390 males and 114 females consisting of an equal number of age- and sex-matched healthy controls and patients with coronary artery disease, were included. The mean age (+/-SD) of the patients and controls were 57.5 +/- 10.6 years and 56.8 +/- 11.0 years, respectively. All the patients underwent coronary angiography; 33, 58, 63, and 98 patients had normal coronaries, single-, two-, or triple-vessel disease, respectively. A third of the patients had suffered from a myocardial infarction. The genotype distribution conforming to Hardy-Weinberg equilibrium was similar for cases and controls. The mean HDL cholesterol increased from B1B1 through B2B2 genotype in controls and sedentary male patients. Self-reported leisure time physical activity, consisting mostly of an hour of morning walk daily, was associated with a rise in mean HDL cholesterol in male controls (33.6 +/- 7.9 mg/dl to 36.2 +/- 8.9 mg/dl, p = 0.037) and patients (32.4 +/- 7.9 mg/dl to 35.7 +/- 11.0 mg/dl; p = 0.018). The exercise-associated rise in HDL cholesterol was most pronounced in controls (32.1 +/- 9.1 mg/dl to 36.8 +/- 9.3 mg/dl, p = 0.05) and male patients (30.5 +/- 7.4 mg/dl to 37.2 +/- 9.7 mg/dl, p = 0.007) with B1B1 rather than B1B2 or B2B2 genotype. The results suggest a possible gene-environment interaction in the regulation of HDL cholesterol that needs to be confirmed in other populations and larger samples to rule out a chance occurrence.