OBJECTIVE Glucocorticoid replacement is the most effective therapy for patients with congenital adrenal hyperplasia (CAH). It has been reported that excessive steroid therapy leads to bone loss and osteoporosis, but it is uncertain whether steroid replacement therapy affects bone turnover and bone mineral density (BMD) in adult patients with CAH. METHODS Case-control study: patients with CAH were compared to normal subjects, individually matched for age and body weight. METHODS Eleven patients, aged 19-65 years, were evaluated in this study. The age at diagnosis of CAH was 0-26 years. Nine patients (6 females and 3 males) were diagnosed with 21-hydroxylase deficiency and 2 male patients with 11-hydroxylase deficiency. 17-Hydroxyprogesterone levels in patients had never been below the reference range in the previous 2 years. These patients were individually matched for sex, age and weight to 11 healthy controls. METHODS Total body, lumbar spine and femoral neck BMD was measured by dual-energy X-ray absorptiometry. Serum bone Gla-protein (BGP, osteocalcin) and bone alkaline phosphatase (BAP) were measured to assess bone formation whereas serum tartrate-resistant acid phosphatase (TRAP) and urinary cross-linked N-telopeptides of type I collagen (NTx) were measured to assess bone resorption. NTx was expressed as a fraction of urinary creatinine excretion (NTx/Cr). Serum dehydro-epiandrosterone sulphate (DHEA-S) and androstenedione levels were measured to assess adrenal androgen status. RESULTS Serum DHEA-S, androstenedione, BGP and BAP and urinary NTx/Cr were decreased in patients when compared with controls (paired t-test, P = 0.005, 0.0003, 0.002, 0.03 and 0.03, respectively). BMD was not significantly decreased in patients. The difference of total body BMD between patients and controls (i.e. BMD in patients minus BMD in controls) was negatively correlated with age. There was no correlation between androgen levels and either BMD or bone turnover. The total dose of steroid taken in the previous 2 years was not correlated with BMD, bone turnover or androgen levels. The was no correlation between BMD adjusted by age and bone turnover or initial age at diagnosis. CONCLUSIONS We conclude that (1) bone turnover is decreased in congenital adrenal hyperplasia, (2) bone mineral density is not decreased in congenital adrenal hyperplasia and (3) patients initially have higher bone mineral density but later have lower bone mineral density than controls, especially in women.