In this study, we tested the hypothesis that there is altered abundance of transcripts of genes coding for the enzymes cytochrome P450 17 alpha-hydroxylase (P450(17 alpha)), cytochrome P450 aromatase (P450arom), and 3 beta-hydroxysteroid dehydrogenase (3 beta-HSD) in follicles of cattle hyperstimulated with eCG compared to FSH. Treatments were initiated on Day 10 of the cycle, and all cows received prostaglandin (PG) on Day 12. In experiment 1, blood samples were taken to determine plasma progesterone and estradiol concentrations during ovarian stimulation. In experiment 2, both ovaries were removed from stimulated cows by colpotomy before (n = 4 cows/treatment) and at 12 (n = 3/treatment) and 24 h (n = 3/treatment) after PG injection, and from nonstimulated controls (n = 4) 72 h after PG. The preovulatory follicle from nonstimulated heifers, and all follicles greater than 3 mm in diameter from superovulated heifers, were isolated and classified as small (3-5 mm), medium (6-9 mm), or large (> 9 mm). Steady-state levels of RNA for 3 beta-HSD, P450(17 alpha), and P450arom genes were determined by Northern analysis in the individual follicles. In experiment 1, stimulation with eCG significantly (p < 0.01) increased plasma progesterone concentrations compared to FSH-stimulated and nonstimulated controls, and increased (p < 0.05) plasma estradiol concentrations compared to FSH-stimulated controls. Stimulation with FSH did not alter progesterone concentrations, but significantly increased plasma estradiol concentrations compared to those of controls. In experiment 2, the number of large follicles increased significantly with time (p < 0.01), but there were no differences between eCG and FSH treatments in size distribution of follicles (p > 0.05). Relative abundance of P450(17 alpha) message (per 20 micrograms RNA) was significantly higher in large and small follicles (p < 0.05) in eCG-treated compared to FSH-treated heifers after PG injection. Analysis within this period revealed significant treatment effects at 12 h but not 24 h after PG injection. The bovine P450arom cDNA hybridized to 3 transcripts: a 6.5-kilobase (kb) polyadenylated transcript, and non-polyadenylated messages of 3.4 and 1.8 kilobases (kb), all of which hybridized with an oligonucleotide probe specific for the heme-binding region. In medium and small follicles, the 6.5-kb and 3.4-kb transcripts were present in similar quantities, and the 1.8-kb transcript was 25% less abundant. In large follicles recovered after luteolysis, the 3.4 and 1.8-kb transcripts were 3- to 4-fold more abundant in eCG-treated compared with FSH-treated and nonstimulated animals (p < 0.05). There were no significant differences between eCG and FSH treatments on steady-state 3 beta-HSD mRNA levels. Levels of 3 beta-HSD and P450(17 alpha) mRNA in large follicles in hyperstimulated heifers were not different from those in preovulatory follicles in nonstimulated cows. We conclude that hyperstimulation with eCG results in greater stimulation of follicular P450(17 alpha) message abundance compared to hyperstimulation with FSH, and that this may contribute to increased follicular estradiol secretion.