In a cross-sectional study using a two-stage probability sample (N = 1,600) of the population of Northern Ireland, there was an inverse association between the highest recorded recent activity and total cholesterol (P < or = 0.01), LDL (P < or = 0.01), triglyceride (P < or = 0.05) and Chol:HDL ratio (P < or = 0.001) in males, and total cholesterol (P < or = 0.001), LDL (P < or = 0.001), and triglyceride (P < or = 0.01) in females; between habitual activity and HDL (P < or = 0.05) in males and total cholesterol (P < or = 0.05) and triglyceride (P < or = 0.01) in females. There was a relationship between the highest recorded activity and apoAI (P < or = 0.01) and apoB (P < or = 0.01) in males and with apoB (P < or = 0.001) in females; between habitual activity and apoAI (P < or = 0.01) and apoAII (P < or = 0.05) in males and apoB (P < or = 0.01) in females; between past activity and Lp(a) in females (P < or = 0.05). After adjustment for possible confounding factors, total cholesterol (P < or = 0.05) and LDL (P < or = 0.05) were unexpectedly higher in males who were active throughout life. Total cholesterol (P < or = 0.05) and LDL (P < or = 0.001) were higher in females with highest recorded activity and triglycerides lower (P < or = 0.05) in those habitually active. An association between highest recorded activity and apoAI (P < or = 0.01), and past activity and apoAI:apoB ratio (P < or = 0.05) was shown in males and in females, after adjustment, and between apoB (P < or = 0.05) and highest recorded activity.