Ten healthy volunteers (six men and four women, aged 22-41 years) were studied in a crossover trial. The study was divided into three one week periods. During each period the subjects either ran on a treadmill, cycled on a bicycle ergometer, or rested in a chair for 1 hour every day. The exercise was performed at two thirds predicted maximum heart rate (equivalent to 50% VO2 max). The sequences were rotated; no studies were performed in the perimenstrual period. Transit was measured by the method of measuring the excretion of a single dose of radio-opaque markers; all stools were collected, weighed, and x rayed after the ingestion of radio-opaque markers. Dietary fibre and fluid intake were measured on the fourth day of each test period by 24 hour record. Lifestyle was otherwise unchanged. Transit time was dramatically accelerated by moderate exercise (both jogging and cycling); however, stool weight, defecation frequency, dietary fibre intake, and fluid intake did not change significantly. Whole gut transit changed from 51.2 hours (95% confidence intervals 41.9 to 60.5) at rest to 36.6 hours (31.6 to 39.2) when riding and 34.0 hours (28.8 to 39.2) when jogging. Riding and running both differed significantly from resting (p less than 0.01); the difference between riding and running was not significant.