Purposes To investigate longitudinally (1) the contribution of morphological covariates to explaining the development of maximum cardiac output (Q max) and maximum arteriovenous oxygen difference (a-vO 2 diff max), (2) sex differences in Q max and a-vO 2 diff max once age, maturity status, and morphological covariates have been controlled for, and, (3) the contribution of concurrent changes in morphological and cardiovascular covariates to explaining the sex-specific development of peak oxygen uptake (V O 2). Methods Fifty-one (32 boys) 11-13-year-olds had their peak V O 2 , maximum heart rate (HR max), Q max, and a-vO 2 diff max determined during treadmill running on three annual occasions. The data were analysed using multilevel allometric modelling. Results There were no sex differences in HR max which was not significantly (p > 0.05) correlated with age, morphological variables, or peak V O 2. The best-fit models for Q max and a-vO 2 diff max were with fat-free mass (FFM) as covariate with age, maturity status, and haemoglobin concentration not significant (p > 0.05). FFM was the dominant influence on the development of peak V O 2. With FFM controlled for, the introduction of either Q max or a-vO 2 diff max to multilevel models of peak V O 2 resulted in significant (p < 0.05) additional contributions to explaining the sex difference. Conclusions (1) With FFM controlled for, there were no sex differences in Q max or a-vO 2 diff max, (2) FFM was the dominant influence on the development of peak V O 2 , and (3) with FFM and either Q max or a-vO 2 diff max controlled for, there remained an unresolved sex difference of ~ 4% in peak V O 2 .