21-Hydroxylase-deficient late-onset adrenal hyperplasia (LOAH) appears to affect 1-6%of hyperandrogenic women. Screening and diagnostic criteria for LOAH have not been well established, as these patients are clinically indistinguishable from other hyperandrogenic women. The following prospective study was undertaken to 1) determine the predictive value of screening hyperandrogenic women for LOAH with a morning follicular phase basal 17-hydroxyprogesterone (17-HP) level and 2) compare the various in vivo estimates of 21-hydroxylase activity after adrenal stimulation for the diagnosis of LOAH. Twenty-one euandrogenic control women (physically normal, without hirsutism, with regular menses, and a negative family history) were studied. The clinical population consisted of 164 consecutive unselected patients seen at the Division of Reproductive Endocrinology and Infertility of Johns Hopkins University School of Medicine between 1983 and 1987 demonstrating hirsutism and/or hyperandrogenic oligomenorrhea. Controls and patients underwent acute adrenal stimulation with 1 mg ACTH-(l-24), administered in the morning to fasting patients in the follicular phase of their menstrual cycle. Blood was sampled before and 30 min after ACTH-(l-24) administration. Steroid RIA determinations were performed for 17-HP, progesterone, testosterone, dehydroepiandrosterone sulfate, androstenedione, FSH, LH, and PRL. Three estimates of 21-hydroxylase activity were studied: the 17-HP level 30 min post-ACTH (17-HP30), the change in 17-HP (Δ17-HPO-30) and the summed rate of change in 17-HP and progesterone ([Δ17-HPo-30) + ΔP0-30]/30 min). The upper 95th percentiles for these estimates of 21-hydroxylase activity in control women were 9.6 nmol/L (316 ng/dL), 8.8 nmol/L (292 ng/dL), and 0.39 nmol/ L-min (13 ng/dL-min), respectively. Thirteen of 164 (7.9%) hyperandrogenic women had at least 1 abnormal 21-hydroxylase measurement. Four of these women (2.4%) had 17-HP measurements 3-to 20-fold above the upper normal 95th percentile (17-HP30 > 36.3 nmol/L or 1200 ng/dL) and were considered as suffering from LOAH. In our population the 3 measures of 21-hydroxylase studied clearly differentiated the LOAH women from all others, although a single 17-HP level 30 min post-ACTH was the simplest and most cost effective. Nine other hyperandrogenic women (5.5%) had at least 1 abnormal 21-hydroxylase measurement less than 3-fold the upper normal 95th percentile value and were designated as having mild 21-hydroxylase deficiency. Basal hormonal values did not serve to distinguish LOAH from other patients, although their mean basal androstenedione and 17-HP levels were significantly higher (P < 0.0002). A basal 17-HP level greater than 6.0 nmol/ L (200 ng/dL) had a positive predictive value for a stimulated 17-HP30 level greater than 36.3 nmol/L (1200 ng/dL), consistent with LOAH, of 80%(95%confidence interval, 28.4-99.5%); the negative predictive value of a basal 17-HP level below 6.0 nmol/L (200 ng/dL) was 100%(95%confidence interval, 97.3-100%). In conclusion, this study confirms the high predictive value of screening for 21-hydroxylase-deficient LOAH in a hyperandrogenic population, with a follicular unsuppressed morning 17-HP level. If the basal 17-HP level is greater than 6.0 nmol/L (200 ng/dL) adrenal stimulation testing should be performed to rule out LOAH, with a 17-HP level 30 min post-ACTH-(l-24) administration above 36.3 nmol/L (1200 ng/dL) consistent with the diagnosis. © 1989 by The Endocrine Society.