Parturition in the ewe is preceded by an increase in the synthesis of prostaglandins (PGs) by gestational tissues. To establish the uterine source of these PGs, placental cotyledons, fetal membranes and maternal uterine tissues were collected from ewes (n=6) at spontaneous parturition. Solubilised tissue extracts were prepared and analysed by Western blots using polyclonal antibodies to PG G/H synthase-1 and -2 (PGHS-1 and PGHS-2). PGHS-1 was expressed by all intrauterine tissues at term labour. Densitometric analysis of Western blot autoradiographs showed that the fetal membranes and maternal cervix contained the largest amounts of PGHS-1. PGHS-1 enzyme content of ovine amnion was significantly greater than that of either chorion or allantois (P<0.05). PGHS-1 protein content of myometrial, endometrial and cotyledonary tissue extracts was minimal. Formation of the PGHS-2 isozyme was confined to placental tissue at term labour. PGHS-2 protein levels in sheep placenta were significantly higher than those of PGHS-1 in all intrauterine tissues examined. This result supports the hypothesis that PGHS-2 is a major contributor to PG formation at term labour. To elucidate the developmental changes in PGHS-1 and PGHS-2 relative to labour onset, an experimental paradigm of glucocorticoid-induced delivery was used. Previous characterisation and validation of this labour model demonstrated that direct, transabdominal, intrafetal injection of the synthetic glucocorticoid betamethasone (5.7 mg in 1 ml aqueous vehicle) on day 131 of gestation induced labour onset in 56.6+/-0.8 h (mean+/-s.e.m.). As the latent period to induced-labour was known, the time course of enzyme formation could be ascertained. Sheep (n=20) were killed by barbiturate injection at various time intervals post-injection (0, 14, 28, 42 and 56 h). Tissue extracts collected at post-mortem examination were prepared and analysed by Western blots. PGHS-2 was induced in ovine cotyledon in a time-dependent fashion following glucocorticoid injection (P<0.05). There was a 12-fold increase in abundance between the time of betamethasone administration (0 h) and established labour (56 h). The PGHS-2 isozyme was not detected in any of the other tissues examined. In contrast, formation of the PGHS-1 isozyme did not change in relation to induced-labour in any of the intrauterine tissues. This finding is consistent with constitutive formation of PGHS-1. Previous studies have demonstrated a rise in PG production in association with glucocorticoid-induced labour and spontaneous delivery. The results of the present study indicate that this rise in PG production is due to increased formation of the PGHS-2 isozyme in ovine cotyledon. PGHS-2 appears to be induced by exogenous glucocorticoid administration and/or the mechanisms controlling ovine parturition. The role of PG formation by the fetal membranes is yet to be elucidated.