This study was designed to determine whether the hypoestrogenic status of 14 amenorrheic athletes was associated with a decrease in regional bone mass relative to that of 14 of their eumenorrheic peers. The two groups of athletes were matched for age, height, weight, sport, and training regimens. Bone mass was measured by dual-photon and single-photon absorptiometry at the lumbar vertebrae (L1 to L4) and at two sites on the radius. Vertebral mineral density was significantly lower in the amenorrheic group (mean, 1.12 g per square centimeter) than in the eumenorrheic group (mean, 1.30 g per square centimeter). There was no significant difference at either radial site. Radioimmunoassay confirmed a lower mean estradiol concentration (amenorrheic group, 38.58 pg per milliliter; eumenorrheic group, 106.99 pg per milliliter) and progesterone peak (amenorrheic group, 1.25 ng per milliliter; eumenorrheic group, 12.75 ng per milliliter) in the amenorrheic women, in four venous samples drawn at seven-day intervals. A three-day dietary history showed no significant differences in nutritional intake, including calcium with and without supplements. The two groups were similar in percentage of body fat, age at menarche, years of athletic participation, and frequency and duration of training but differed in number of miles run per week (amenorrheic group, 41.8 miles [67.3 km]; eumenorrheic group, 24.9 miles [40.1 km]). We conclude that the amenorrhea that is observed in female athletes may be accompanied by a decrease in mineral density of the lumbar vertebrae.