A shift in progesterone-to-estradiol balance to estradiol dominance is assumed to be a prerequisite for regular uterine contractions. To antagonize this effect in premature labor 24 consecutive women were treated with intravenous cortisol for 3 days and with weekly intramuscular injections of 17 alpha-hydroxyprogesterone caproate (17 OHP-C). Twenty-four similar patients treated with ritodrine served as a reference group. The delivery was postponed by at least 1 week in 21 patients (87.5%) in the steroid treatment group and in 18 patients (75%) in the ritodrine group. The premature labor lasted for 5.1 +/- 0.4 hours (mean +/- SEM) with steroid therapy and for 2.2 +/- 0.3 hours with ritodrine. In singleton pregnancies the gestational length and birth weight of the newborn infants were greater in the steroid treatment group (N = 23, 39.1 +/- 0.3 weeks, 3,460 +/- 119 gm) than in the ritodrine group (N = 24, 37.7 +/- 0.4 weeks, 3,106 +/- 118 gm). Steroid treatment suppressed serum estradiol concentrations (maximally by 60%) and, to a lesser extent, testosterone, estriol, and progresterone levels (maximally by 30%).