In a comparative trial, ethanol effectively arrested preterm labor for 48 h or more in 32% of cases, a .beta.-adrenergic agent (salbutamol) in 60% of cases (not statistically significant), and a combination of ethanol and indomethacin in 70% of cases (statistically significant, P < 0.5). Labor was delayed for 14 days or more in 36, 60, and 50%, respectively (not statistically significant). The numbers studied were small, and tests of statistical significance were of doubtful value. Salbutamol was more acceptable to patients and the staff than ethanol. The trial was suspended and eventually abandoned due to reports of prostaglandin synthetase inhibitors causing premature closure of the ductus arteriosus. In this series no problems were encountered with the use of indomethacin.