Physical fitness and leisure time physical activity are strongly correlated, and both are inversely correlated with risk of ischaemic heart disease. Does this mean, however, that a very fit man has a lower risk of ischaemic heart disease (IHD), even if he is inactive? And does it also mean that an unfit, but active man, does not have a lower risk of IHD than an unfit, inactive man? In the Copenhagen Male Study, we analysed the joint effect of fitness and leisure time activity. In 1970/71, 4999 men aged 40-59 years, were classified according to level of physical fitness, i.e. indirectly measured maximal oxygen uptake, and physical activity, and their mortality was recorded over the following 17 years. In sedentary men, fitness was no predictor of future risk of IHD whatsoever. Age-adjusted baseline values were similar in later IHD cases and survivors (32.3 and 32.1 ml O2 kg-1 min-1, respectively: P = 0.91). In medium or highly active men, however, fitness was a strong predictor. The corresponding fitness values were 33.1 and 34.8 ml O2 kg-1 min-1 (P < 0.001). The least fit (two least fit quintiles) physically active men had a lower IHD mortality rate (6%) than the least fit sedentary men (10%). Adjusted for age, social class and smoking in a multiple logistic regression equation, this was estimated to an RR (95% C.I.) of 1.67 (1.06-2.64) (P = 0.027). The two major new findings of this study were (a) that being very fit, provides no protection against IHD--nor all-cause mortality--in sedentary men, and (b) that unfit but sedentary men have a higher risk of IHD than unfit but active men, i.e. those performing light physical activity for at least 4 h per week.