The purpose of this study was to compare three weight-training methods to measures of high-intensity exercise endurance. Young male subjects were assigned to three groups using the same exercises three days per week. The initial age, height and body mass of the subjects were (mean ± standard deviation): Group N (n = 8, 20.0 ± 2.0 years, 181.0 ± 5.7 cm, 74.5 ± 11.5 kg); Group P (n = 9, 19.3 ± 1.2 years, 179.5 ± 3.0 cm, 72.4 ± 6.4 kg); Group H (n = 10, 20.6 ± 5.0 years, 184.2 ± 6.4 cm, 80.2 ± 11.2 kg). Parallel squats were performed two days per week and 1/4 squats one day per week. Group N used one set of approximately 12 repetitions to failure. Group P used three sets of 10 for two weeks, three sets of five for three weeks and three sets of three for two weeks. Group H used three sets of 10 for the entire seven weeks. Groups P and H used light and moderate warm-up sets before exercise with the three target sets. Body mass was measured on a medical scale. Cycle endurance time (CT) was measured with the subjects riding (60 rpm) at 30 watts (two minutes), 120 watts (two minutes) and 265 watts to exhaustion. Squatting (top of the thigh parallel) endurance measurements began at 60 kg at a cadence of 1 squat per six seconds. Bar mass was raised by 2.5 kg each minute until exhaustion. Maximum mass lifted (MM), total repetitions (TR) and load (repetitions x mass) (L) were calculated. Measures were made at the beginning (T1) and after seven weeks of training (T2). No differences were found between groups; however, within- group analysis showed significant (p < 0.05) increases over time for both P and H, but not for N on all measures. These data suggest that one set to failure does not increase high-intensity exercise endurance as effectively as the use of multiple sets of weight training.