We report 98 children who have reached final adult height in a long-term trial of LHRH agonist treatment. These children were 5.3 +/- 2.1 yr old at the start of treatment and were treated with either deslorelin (4 microg/kg.d sc) or histrelin (4-10 microg/kg.d) for an average of 6.1 +/- 2.5 yr. Final height averaged 159.8 +/- 7.6 cm in the 80 girls, which was significantly greater than pretreatment predicted height (149.3 +/- 9.6 cm) but still significantly less than midparental height (MPH) (163.7 +/- 5.6). Final height averaged 171.1 +/- 8.7 cm in the 18 boys, which was significantly greater than pretreatment predicted height (156.1 +/- 14.2 cm) but still significantly less than MPH (178.3 +/- 5.2 cm). However, the average adult height of the 54 children who had less than a 2-yr delay in the onset of treatment was not significantly different from their MPH, and 21 children exceeded MPH. Final height SD score correlated positively with duration of treatment (P < 0.01), midparental height (P < 0.001), predicted height at the start of treatment (P < 0.001), and growth velocity during the last year of treatment (P < 0.001) and correlated inversely with delay in the onset of treatment (P < 0.001), age at the start of treatment (P < 0.001), bone age at the start of treatment (P < 0.001), bone age at the end of treatment (P < 0.001), breast stage at the start of treatment (P = 0.02), and bone age minus chronological age at the start of treatment (P = 0.001). We conclude that LHRH agonist treatment improves the final height for children with rapidly progressing precocious puberty treated before the age of 8 yr for girls or 9 yr for boys. Less delay in the onset of treatment, longer duration of treatment, and lower chronological and bone age at the onset of treatment all lead to greater final height. All children with onset of pubertal symptoms before age 8 in girls and age 9 in boys should be evaluated for possible treatment. Treatment is appropriate in children with rapidly progressing puberty, accelerated bone maturation, and compromise of adult height prediction, regardless of bone age or chronological age at time of evaluation. However, once treatment is considered appropriate, it should be initiated quickly, because longer delays lead to shorter final height. In addition, the longer the treatment is continued, the greater is the final height outcome.