Abstract This study attempted to validate an anthropometric equation for predicting age at peak height velocity (APHV) in 193 Polish boys followed longitudinally 8-18 years (1961-1972). Actual APHV was derived with Preece-Baines Model 1. Predicted APHV was estimated at each observation using chronological age (CA), stature, mass, sitting height and estimated leg length. Mean predicted APHV increased from 8 to 18 years. Actual APHV was underestimated at younger ages and overestimated at older ages. Mean differences between predicted and actual APHV were reasonably stable between 13 and 15 years. Predicted APHV underestimated actual APHV 3 years before, was almost identical with actual age 2 years before, and then overestimated actual age through 3 years after PHV. Predicted APHV did not differ among boys of contrasting maturity status 8-11 years, but diverged among groups 12-15 years. In conclusion, predicted APHV is influenced by CA and by early and late timing of actual PHV. Predicted APHV has applicability among average maturing boys 12-16 years in contrast to late and early maturing boys. Dependence upon age and individual differences in actual APHV limits utility of predicted APHV in research with male youth athletes and in talent programmes.