Symptom-limited, graded exercise treadmill testing was performed by 4,968 white and black adults, ages 18-30 yr, during the baseline examination for the Coronary Artery Risk Development in Young Adults (CARDIA) study. Compared with nonsmokers, the mean exercise test duration of smokers was 29-64 s shorter depending on race/gender group (all P < 0.001), but mean duration to heart rate 130 (beats.min-1) ranged from 20-50 s longer (P < 0.05). In each race/gender group, test duration to heart rates up to 150 was 15-35 s longer (P < 0.05) in smokers than in nonsmokers after adjustment for age, sum of skinfolds, hemoglobin, and physical activity score. The mean maximum heart rate was lower in smokers than in nonsmokers (difference ranging from 6.7 beats.min-1 in white men to 11.2 beats.min-1 lower in black women, P < 0.001), although maximum rating of perceived exertion was nearly identical in smokers and nonsmokers. Chronic smoking appears to blunt the heart rate response to exercise, so that exercise duration to submaximal heart rates is increased even though maximal performance is impaired. This may result from downloading of beta-receptors caused by smoking. Smoking status should be considered in the evaluation of physical fitness data utilizing submaximal test protocols, or else the fitness of smokers relative to nonsmokers is likely to be overestimated.