A radioimmunoassay for dihydrotestosterone (DHT) in plasma was developed using an antiserum raised against testosterone-3-oxime-bovine\x=req-\ serum-albumin. After extraction of 1 ml male plasma with diethylether, DHT was separated from testosterone (T) by thin-layer chromatography. A dextran-charcoal-suspension was used for the separation of bound and free ligand. The inter-assay variability was 10.4 % (C. V.) and the detection limit 1.77 ng/100 ml. The accuracy of the method as determined by mass recoveries and the specificity were shown to be satisfactory. Normal values were obtained in 45 young to middle-aged (22\p=n-\61years) and 37 old (68\p=n-\93years) men. The median and the 95 percentiles were 20.5\p=n-\51.9\p=n-\76.3 (ng/100 ml) and 19.5\p=n-\50.9\p=n-\101.5 (ng/100 ml) respectively. While DHT did not change in old age T fell by 20.6%. DHT and T showed a significant correlation: rS = 0.426, P < 0.01 (young men), rS = 0.752, P < 0.001 (old men). After 3 daily im injections of 5000 IU human chorionic gonadotrophin (HCG), DHT increased 1.50 times (range: 1.15\p=n-\2.09,n = 12), T 1.86 times (range: 1.20\p=n-\2.91,n = 12). After 4 daily administrations of 40 mg fluoxymesterone DHT fell to 29.6% of the control level (range: 16.0\p=n-\48.2%, n = 12). Blood samples were obtained from a 24 year old man every 15 min for 24 h. A close parallelism was observed between the concentrations of DHT and T in the plasma.In recent years a good deal of evidence has been accumulated that dihydro¬ testosterone (DHT), the main metabolite of testosterone (T) in male target organs, may be the active hormone with regard to many but not all androgen