Twenty-four men and 26 women (25 +/- 5 years) participating in 10 weeks (n = 27) and 18 weeks (n = 23) of variable resistance strength training programs were recruited to complete 12 weeks of reduced training. Training consisted of one set of 7-10 bilateral knee extensions performed to volitional failure. Prior to the reduced training phase of the project, the subjects were training either 2 days.week-1 (n = 23) or 3 days.week-1 (n = 18). The subjects who trained 3 days.week-1 reduced training frequency to 2 days.week-1 (n = 9), 1 day.week-1 (n = 7), or 0 days.week-1 (n = 2). The subjects who trained 2 days.week-1 reduced training frequency to 1 day.week-1 (n = 12) or 0 days.week-1 (n = 11). Nine subjects served as controls and did not train. Isometric knee extension strength was assessed at 9, 20, 35, 50, 65, 80, 95, and 110 degrees of knee flexion on two separate occasions prior to and immediately post-training and following reduced training. After training, mean relative increases in peak isometric knee extension strength and dynamic training weight were 21.4% +/- 17.5% (P less than or equal to 0.01) and 49.5% +/- 14.7% (P less than or equal to 0.01), respectively. The subjects who stopped training (0 days.week-1) lost 68% (P less than or equal to 0.01) of the isometric strength gained during training.(ABSTRACT TRUNCATED AT 250 WORDS)