Nine male pairs of monozygotic twins aged 11-14 years, height 147 (7.6) cm and body mass 39.7 (9.6) kg, participated in this study. Twin zygocity was tested using morphological, dermatoglyphic and hematologic methods, and Tanner's five stages were used for the evaluation of biological maturation. One twin from each pair undertook training for 6 months, three times a week, with running at 85-120% of the lactate anaerobic threshold (LT). Anthropometrics, determination of maximum O(2) uptake (.VO(2max)), LT and maximal blood lactate concentration ([La](max)) was carried out before, during and after training. No significant difference existed between the trained twins and their untrained brothers before training. After training, the trained twins increased their .VO(2max), (per kg body mass) by 10.6% and their LT by 18.2% (P<0.01), reaching values that differed significantly from those of their untrained brothers [57.5 (3.6) ml.kg(-1).min(-1) vs 55.4 (3.3) ml.kg(-1).min(-1) and 13.4 (1.1) km.h(-1) vs 12.7 (1.1) km.h(-1), respectively]. In addition, in the trained twins relative body fat was reduced (P<0.05) from 17.8 to 16.2% and their somatotype altered significantly (decrease of endomorphy and mesomorphy and increase of ectomorphy), while in the untrained twins there was no change in these parameters. Both groups of twins significantly increased their absolute .VO(2max) after the 6 months of training, the trained by 14,9% [from 2.08 (0.43) to 2.37 (0.45) l.min(-1)] and the untrained by 10.5% [from 2.10 (0.41) to 2.32 (0.47) l.min(-1)], but no difference was registered between them. A comparison of the intrapair changes in .VO(2max) of prepubertal and pubertal twins showed an influence of training in the prepubertal (19.3% vs 5.2%) but not in the pubertal twins (12.7% vs 13.1%). Using analysis of variance, the relative importance of training, heredity and their interaction was evaluated to be 20%, 70% and 10%, respectively, for the change in body fat, 35%, 45% and 20%, respectively, for the change in relative .VO(2max) and 25-30%, 50-60% and 15-20%, respectively, for the change in LT. In conclusion, training during pubertal growth can favour aerobic power (depending on body composition) as well as aerobic capacity, but it has no effect on absolute .VO(2max). Genetic control seems to have a strong effect on the extent of adaptations, and the genotype-training interaction explains a small, but prominent part of them.