Ninety-seven patients with definite or classic rheumatoid arthritis (RA); 54 patients (19 premenopausal women, 25 postmenopausal women and 10 men) were treated with low-dose glucocorticoids for at least 12 mo. (mean dose < 10 mg/day). The remaining 43 patients (15 premenopausal women, 17 postmenopausal women and 11 men) were treated with penicillamine and served as a patient control group. The distal forearm bone mineral content (BMC) was measured in all patients by single photon absorptiometry using 125I. The total body bone mineral (TBBM) was measured in 61 patients by dual photon absorptiometry using 153Gd. Compared with normal controls, both treatment groups had significantly decreased BMC and TBBM (0.01 .+-. P < 0.001). When the patients were stratified according to pre- and postmenopausal state, significantly lower BMC and TBBM values were found in the premenopausal glucocorticoid-treated women than in pencillamine-treated women. No differences in BMC and TBBM values were found in the corresponding postmenopausal groups. In the premenopausal women treated with glucocorticoids, the duration of treatment and cumulative dose correlated with BMC. No such correlations were found in the postmenopausal women. RA is associated with loss of bone mass; systemic glucocorticoid treatment further aggravates bone loss. In postmenopausal RA patients, the bone loss resulting from menopause and from the disease itself is not accelerated by low-dose glucocorticoids. In premenopausal RA patients, the bone mass is significantly affected by glucocorticoid treatment. These factors should be considered when prescribing glucocorticoids, in order to minimize bone loss.