Three antiprogestogens of the RU 38.486, ZK 98.734, and ZK 98.299, were studied at different stages of pregnancy in the guinea pig. Treatment starting on postconception day 4 completely prevented nidation; all three compounds had comparable inhibitory potency. Treatment after nidation, starting on postconception day 8, induced decidual collapse and bleeding, but embryonic tissue was retained in nidation sites. In contrast to results in animals in nonfertile cycles, luteolysis was not induced, indicating that antiprogestogens lack the ability to induce uterine prostaglandin synthesis/liberation. On postconception day 43, RU 38.486 showed marginal abortifacient activity. The other compounds induced expulsion more rapidly and at a higher rate. The comparatively pronounced antiglucocorticoid activity of RU 38.486 may account for this difference. With RU 38.486, a high level of uterine prostaglandin sensitivity and a cervical ripening were induced consistently and fast; spontaneous labor, on the other hand, occurred after several days, if at all. Complete uterine evacuation was induced within hours by otherwise inactive doses of sulprostone in various combinations with ZK 98.299 RU 38.486 but surprisingly not with ZK 98.734. A single dose of ZK 98.299 induced an approximately thirtyfold increase in uterine prostaglandin sensitivity within 24 hours, exceeding that present before term, but did not induce spontaneous labor. This is evidence that endogenous prostaglandins were not activated, analogous to perinidation stages. Observation of antiluteolytic activity of antiprogestogens in nonpregnant animals is considered of major theoretical importance in this context. It seems that inhibition of progesterone leads to suppressed uterine prostaglandin liberation. The same effect in pregnancy could explain the inability of the uterus to expel a seriously compromised conceptus. In conclusion, we suggest that progesterone is a stimulator rather than a depressor of uterine prostaglandins in the late luteal phase and pregnancy. The ability of the conceptus to neutralize this stimulatory action of progesterone is considered to be essential for the rescue of the corpus luteum and uterine motor quiescence in the guinea pig. The clinical significance of these findings is that the high frequency of incomplete abortions and protracted, sometimes heavy bleeding in pregnant women treated with RU 38.486 may reflect decidual compromise and simultaneous uterine prostaglandin deficiency, as found in our animal model after progesterone blockage.