Term pregnant patients undergoing preinduction cervical softening were randomized to receive no treatment (controls) or 0.5 mg PGE2-triacetin gel (Prepidil gel) endocervically. Plasma samples containing PGEM collected 4 hrs post-treatment and converted to the stable bicyclo degradation product (bicyclo-PGEM) were assayed by RIA. Positive clinical effect (responders) during 12 hrs after treatment (Bishop score increase greater than or equal to 3, in labor or delivered) were assessed. All evaluations were blind. In nonresponders (n = 35), the means of the bicyclo-PGEM variables (mean, maximum, area under the curve) were all about 18% higher in gel-treated patients (n = 6) than controls (n.s.). In responders (n = 38), the variables were all about 80% higher in gel-treated women (n = 32) than controls (p less than .01). In controls (n = 35), the responders (n = 6) had 50% higher levels than nonresponders (n.s.). In the gel-treated women (n V 38), responders (n = 32) had about 140% higher levels than nonresponders (less than .01). The results suggest that both exogenous and endogenous bicyclo-PGEM were measured. Differences in pairwise comparisons suggest that there may be substantially less exogenous bicyclo-PGEM in the gel nonresponders than in gel responders or substantially more endogenous bicyclo-PGEM in gel responders than in control-responders.