The physiology of marathon running has been extensively studied both in the laboratory and in the field, but these investigations have been confined to elite competitors. In the present study 28 competitors who took part in a marathon race (42.2 km) have been studied; 18 male subjects recorded times from 2 h 19 min 58 s to 4 h 53 min 23 s; 10 female subjects recorded times between 2 h 53 min 4 s and 5 h 16 min 1 s. Subjects visited the laboratory 2-3 weeks after the race and ran on a motor driven treadmill at a series of speeds and inclines; oxygen uptake (VO2) was measured during running at average marathon racing pace. Maximum oxygen uptake (VO2 max) was measured during uphill running. For both males (r = 0.88) and females (r = 0.63), linear relationships were found to exist between marathon performance and aerobic capacity. Similarly, the fraction of VO2 max which was sustained throughout the race was significantly correlated with performance for both male (r = 0.74) and female (r = 0.73) runners. The fastest runners were running at a speed requiring approximately 75% of VO2 max; for the slowest runners, the work load corresponded to approximately 60% of VO2 max. Correction of these estimates for the additional effort involved in overcoming air resistance, and in running on uneven terrain will substantially increase the oxygen requirement for the faster runners, while having a much smaller effect on the work rate of the slowest competitors. Five minutes of treadmill running at average racing pace at zero gradient did not result in marked elevation of the blood lactate concentration in any of the subjects.