Previous studies from our laboratory have suggested that post-receptor events at the level of β-adrenergic receptor—adenylate cyclase interaction could be altered in myometrium by steroid hormones or pregnancy. In this study, we have addressed this question by performing a direct evaluation of rat myometrial Gs proteins at various stages of pregnancy or 24 h after administration of progesterone. In the 50 000 g myometrial plasma membrane fraction, in the presence of 32 P-labelled NAD, cholera toxin ribosylated three predominant proteins with apparent molecular masses of 42, 47 and 55 kDa. Western blot analysis using the RM/1 antibody recognized the 42 and 47 kDa cholera toxin ADP-ribosylated bands but not the 55 kDa band. Thus, the 42 and 47 kDa immunoreactive bands were interpreted as being the small (Gsα-S) and large (Gsα-L) forms of Gs respectively. With a more purified myometrial plasma membrane fraction (105 000 g ) an additional minor band of 44 kDa could be observed with both techniques. Treatment of late pregnant rats with 5 mg progesterone resulted in a significant increase in both Gsα subunits: +25% and +30% after ADP-ribosylation, +50% and +60% after Western blot analysis for Gsα-L and Gsα-S respectively. Pretreatment with the antiprogestin RU 486 completely suppressed the effect of progesterone, suggesting that the expression of Gsα subunits may be under the control of progesterone. However, changes in the myometrial content of Gs in progesterone-treated rats were not associated with concomitant variations in the steady-state levels of mRNA as demonstrated by Northern blot analysis. These data suggest a post-translational regulation of Gs expression by progesterone. Amounts of ADP-ribosylated Gs showed characteristic changes during the course of pregnancy with a fourfold or threefold increase (P<0·05) on day 15 versus day 12 or delivery respectively. During pregnancy, or after progesterone administration, myometrial alterations of Gs strongly correlated (r=0·913, P<0·01) with the cholera toxin-stimulated adenylate cyclase activity. These findings provide evidence that changes in myometrial amounts of functional Gs i) are controlled by the hormonal status of pregnancy and progesterone and ii) play an important role in the transduction pattern of adenylate cyclase activity during the course of pregnancy.