This study compares the effects of 3 weight-training programs on the 1 repetition maximum (1RM) squat (SQ). Subjects were 21 college-age men. The criteria for subject inclusion was initial 1RM >110 kg and >1.3 x body mass and the ability to complete >80% of the programmed repetitions. The groups were Group 1 (n = 5) 5 x 6RM, Group 2 (n = 9) stepwise periodized model, and Group 3 (n = 7) overreaching periodized model. Groups 1 and 2 were equalized on programmed repetitions (720 and 732), and Group 3 was programmed at 18 and 19.4% fewer repetitions (590). Actual repetitions achieved for Groups 1-3 were 619, 629, and 529, respectively. The 1RM squat was measured before and after 12 weeks. Within-group analysis showed that only Groups 2 and 3 increased significantly (p < 0.05) in the 1RM (kg +/- SD): Group 1, 141.4 +/- 28.1-155.4 +/- 23.7; Group 2, 124.8 +/- 12.0-143.4 +/- 12.1; and Group 3, 132.8 +/- 17.0-153.3 +/- 19.3. Derived variables were squat (SQ) x body mass(-1) and SQ gain score x Sinclair coefficient (the method of obviating differences in body mass). Percent differences between groups for Groups 1 and 2 were SQ = 33, SQ x body mass = 53, and Sinclair formula = 33. For Group 3, SQ = 46, SQ x body mass = 67 and Sinclair formula = 109. These data indicate that periodized models increased the 1RM squat to a greater extent than a constant repetition scheme, even when the repetitions were equalized (Group 1 vs. Group 2) or when the repetitions were substantially fewer (Group 1 vs. Group 3).