The incidence of cardiac arrhythmias observed during maximal treadmill exercise testing was studied in 650 men aged 25 to 54 years. In 561 there was no clinical evidence of cardiovascular disease; 89 were classified as having definite or suspected cardiovascular disease. The patients were divided into 3 age groups—25 to 34, 35 to 44 and 45 to 54 years—to define any age-related differences in the incidence of arrhythmias during exercise. Single or consecutive ventricular premature complexes were observed in 31 percent of the 25 to 34 year olds, 38 percent of the 35 to 44 year olds and 49 percent of the 45 to 54 year olds; the incidence of supraventricular premature complexes was 7, 10 and 14 percent, respectively. The incidence of both ventricular and supraventricular complexes increased with age. For any given age group the incidence of ventricular premature complexes was greater in patients with definite or suspected cardiovascular disease. These patients were more prone to demonstrate frequent ventricular premature complexes and had a higher incidence of multifocal ventricular premature beats and ventricular tachycardia. The ventricular premature complexes were also more likely to appear at lower heart rates during exercise in patients with cardiovascular disease. However, the appearance of unifocal ventricular premature complexes during maximal or near maximal exercise testing should not be equated with the presence of clinically significant cardiac disease.