5 alpha-Androstane-3 alpha, 17 beta-diol (A3 alpha diol) is a potent androgen, and is an end product of testosterone. Many authors measured A3 alpha diol levels in human plasma by various methods, but the levels of this steroid were very dissimilar. In order to validate such values, it was measured by gas chromatography-selected ion monitoring (GC-SIM) in this study. A3 alpha diol, 5 alpha-Androstane-3 beta, 17 beta-diol (A3 beta diol) and Testosterone (T) in human peripheral serum were measured by GC-SIM at the same time. The TFA-derivatives of these compounds were analyzed after purification of the serum extract by Sephadex LH-20 microcolumn chromatography. The sensitivity was good: (16.7 pg/ml: A3 alpha diol, 26.7 pg/ml: A3 beta diol). The precision (CV = 2.75%: A3 alpha diol, 3.11%: A3 beta diol) and the accuracy were better than ever reported. Serum A3 alpha diol was measured in 131 healthy men aged 15-81 years and 5 healthy women aged 25-60 years. There were remarkable differences between individuals in the serum levels of A3 alpha diol, but the levels in male serum (greater than 20y) showed a significant negative correlation with age (r = -0.560, p less than 0.01). When these healthy men were classified into three age groups of 20-39, 40-59 and 60-79 years, the values (mean +/- SD) for serum A3 alpha diol were 189.3 +/- 77.7 (n = 20), 127.9 +/- 59.5 (n = 28), and 94.9 +/- 52.9 (n = 73) pg/ml, respectively. There were significant differences between the levels of this steroid in all age groups (p less than 0.01). There was a weak but significant correlation between serum A3 alpha diol and T levels (r = 0.3235, p less than 0.01) in healthy men (25-77 years, n = 77). Determination of serum A3 alpha diol was influenced by age. The number of samples strongly influenced the decision of mean value of A3 alpha diol levels. These results suggested that these factors had to be made obvious when this steroid was studied.