A sensitive and reliable radioimmunoassay (RIA) for the measurement of norethindrone (NET) in serum has been established employing anti-11 alpha-hydroxynorethindrone 11-hemisuccinyl-bovine serum albumin serum in conjunction with norethindrone-3-(0-carboxymethyl) oximino-[125I]-iodohistamine. Of a number of ring A reduced NET metabolites, only 17 beta-hydroxy-17 alpha-ethinyl-5 beta-estran-3-one (43%) and 17 alpha-ethinyl-5 alpha-estrane-3 beta, 17 beta-diol (15.7%) cross-reacted appreciably in this RIA. Ethinyl estradiol (EE2) and mestranol (MEE2) exhibited cross-reactions of only 1.1 and 0.4%, respectively. Serum NET levels were measured in four groups of 3 women, each ingesting either 1 mg NET plus 0.05 mg MEE2 (Norinyl 1 + 50 or Ortho Novum 1/50), 0.5 mg NET plus 0.035 mg EE2 (Brevicon) or only 0.35 mg NET (Micronor) daily for 5 consecutive days. Peak serum NET levels were observed within 1/2 to 4 hours after oral intake and fell precipitously thereafter. After reaching a maximum, serum NET concentrations declined in a manner consistent with at least two disposition phases. The average half-life for the first disposition phase was 2.3, 3.4, 3.9 and 4.4 hours in subjects ingesting Norinyl 1 + 50, Ortho Novum 1/50, Brevicon and Micronor, respectively. Peak and 3-hour post-ingestion serum NET concentrations were dose-related but showed considerable subject-to-subject variations. Following discontinuation of tablet intake, serum NET levels remained detectable (greater than 0.05 ng/ml) for at least 5 days in all 3 women who had taken Ortho Novum 1/50, but in none of the other 9 volunteers. These results suggest that different preparations of identical doses and combinations of oral contraceptive steroids may yield different serum NET profiles. However, due to considerable subject-to-subject variations, larger numbers of subjects are required for a conclusive investigation.