The influences of gender and sexual maturation on the peak VO2 of 12-yr olds were examined. Subjects were 106 boys and 106 girls, ages 12.2 +/- 0.4 yr. The sexual maturity of 93 boys and 83 girls was classified according to Tanner's indices of pubic hair. No significant gender differences (P > 0.05) were detected in age, stature, or hemoglobin concentration. Peak VO2 was determined on a treadmill and boys' peak VO2 was significantly higher (P < 0.01) than girls' whether expressed in L x min(-1) (2.10 +/- 0.34 vs 1.92 +/- 0.28) or mL x kg(-1) x min(-1) (52 +/- 6 vs 44 +/- 5). With body mass controlled for using log-linear ANCOVA, the gender difference decreased from 18.2 to 17.1% but remained significant (P < 0.01). For peak VO2 (L x min[-1]) ANOVA revealed no significant interaction (P > 0.05) but significant (P < 0.01) main effects for both gender and maturation. For peak VO2 in ratio with body mass (mL x kg(-1) x min[-1]), ANOVA detected no significant interaction (P > 0.05) or significant main effect (P > 0.05) for maturation although the main effect for gender was significant (P < 0.01). Analysis of peak VO2 with body mass controlled for using log-linear ANCOVA revealed no significant interaction (P > 0.05) but main effects (P < 0.01) for both gender and maturation. Thus, gender differences, which are not simply explained by differences in body size or hemoglobin concentration, have been demonstrated in the peak VO2 of 12-yr olds. A log-linear scaling model has identified in both boys and girls a significant influence of maturation on peak VO2 independent of body mass. This effect may have been masked in previous studies by the inappropriate use of peak VO2 in ratio with body mass.