Bone density was studied in intercollegiate athletes and older athletic women. Single-photon densitometry was used to assess bone density parameters at a new distal radial site, the midradius, and the first metatarsus. Dual-photon densitometry assessed bone density of the lumbar spine. Eleven intercollegiate tennis players, 23 swimmers, and 86 older "athletic" women from 23 to 75 years of age were compared with age-matched nonathletic controls. "Athletic" describes adult women who exercised at least three times per week, 8 or more months of the year, for a minimum of 3 years. The radius and metatarsus bone content of intercollegiate athletes was significantly above control values. Lumbar spine density was significantly higher only in tennis players. Mean bone density values for adult "athletic" women were also significantly greater than in age-matched controls. In the oldest athletic group (55-75 years of age) bone measurement values in radius and lumbar spine were in the same range as for younger "athletic" women. In contrast, after 50 years of age, these values in the control population decreased by 0.7%/year. Therefore the largest variance (increase) from age-matched controls occurred in the oldest "athletic" group. Also, we have established a distal radial density value (using our modified site) below which we consider women "at risk" and recommend further bone health evaluation. Only two adult athletic women greater than 55 years of age fell into this category. It is concluded from this cross-sectional study that a regularly maintained athletic program for adult women may reduce the rate of "normal" bone mass loss accompanying age, particularly postmenopausally.(ABSTRACT TRUNCATED AT 250 WORDS)