We administered the aromatase inhibitor fadrozole to 16 girls with gonadotropin-independent precocious puberty due to the McCune-Albright syndrome. The girls' ages ranged from 3.2-9.7 yr, and their bone ages ranged from 5.75-14.25 yr. After baseline evaluations, fadrozole was started at a dose of 240 microg/kg.d (equivalent to the dose recommended for therapy of estrogen-dependent breast cancer) for 12-21 months and increased to 480 microg/kg.d for an additional 12 months in 10 girls. During treatment, seven girls had evidence of central precocious puberty; hence, the GnRH agonist deslorelin (4 microg/kg.d sc) was added to their regimen. One girl was on a long-acting GnRH agonist from the start of treatment. Patients were evaluated at 2-6-month intervals throughout treatment. After the first 6-12 months of treatment, fadrozole showed some benefits in 10 girls, including decrease in frequency of menses and/or rates of linear growth and bone maturation; however, fadrozole had no significant benefit in the group as a whole. The seven girls with evidence of central precocious puberty had no slowing in the progression of their puberty during the combined fadrozole and GnRH analog treatment. Adverse effects of fadrozole included inhibition of cortisol and aldosterone biosynthesis at the dose of 480 microg/kg.d, without clinical evidence of adrenal insufficiency. In addition, three patients complained of nonspecific abdominal pain during fadrozole treatment. In one patient, this resolved with a reduction in dose from 480 to 240 microg/kg.d; in two patients, it resolved spontaneously. One girl had muscle weakness and constipation on the 480 microg/kg.d. This resolved after discontinuation of the drug. We conclude that fadrozole is not sufficiently potent to block estrogen synthesis in most girls with gonadotropin-independent precocious puberty due to the McCune-Albright syndrome and may impair the adrenocortical stress response.