Relationships between motor performance, as measured by various fitness tests, and age at peak height velocity have been studied in a sample of 173 Flemish boys, measured yearly between +/- 13 and +/- 18 years and again as adults at 30 years of age. In addition to correlation studies, comparisons were made between boys with an early, average and late age at peak height velocity. To summarize the successive measurements during adolescence, a longitudinal principal component analysis was carried out. The first component can be interpreted as an average percentile level component. During adolescence, three performance tasks, namely speed of limb movement, explosive strength and static strength, are negatively related to age at peak height velocity; thus early maturers performed significantly better than late maturers. However, between late adolescence and adulthood, a cross-over of the average distance curves between 18 and 30 years of age was noted for almost all motor tasks. The late maturers not only caught up the early maturers, but there were significant differences for explosive strength and functional strength in favour of late maturers. In order to predict performance in adulthood from measures during adolescence, the following hypothesis is suggested: the best results at adulthood are obtained by those men who were already good performers during adolescence and who were late maturers, while the worst results are obtained by poor performers during adolescence who were early maturers.