SummaryWe evaluated the accuracy of height predictions based on the tables of Bayley and Pinneau (2) in 43 boys with short stature. Sixteen boys were treated with androgens and 27 received no treatment. In 17 boys whose bone ages were within normal limits, and who received no treatment, the mean ± SE predicted height of 164.9 ± 1.5 cm was not significantly different from the mean adult height (166.5 ± 1. 5 cm). The predicted height exceeded the actual adult height by more than 5.1 cm in only one instance [5.1 cm is the degree of accuracy reported by Bayley and Pinneau (2)]. In 10 boys, whose bone ages were severely delayed (more than 2 SD below their chronologic age) and also were not treated, predicted height overestimated adult height by more than 5.1 cm in five of them. This difference was statistically significant (P<0.05). In five boys with normal bone ages, androgen therapy had no significant effect on either predicted height (168.1 ± 4.1 before, 166.8 ± 4.4 cm after) or actual adult height (166.5 ± 4.1 cm). The 11 boys with severely delayed bone ages had a significant increase in predicted height during androgen therapy (165.4 ± 1.5 to 169.8 ± 1.7 cm, P<0.01), but actual adult height (162.4 ± 2. 4 cm) was not significantly greater than pretreatment predicted height. Further, the number of boys whose predicted height exceeded their adult height by 5.1 cm was not significantly different in treated (4/11) or untreated (5/10) boys. The Bayley-Pinneau method of predicting adult height markedly overestimates adult height in about one-half of boys who have a bone age delay of more than 2 SD. Short-term therapy with androgens does not alter this outcome.