A preterm birth prevention program consisting of risk scoring, intensive weekly observation including cervical examinations, and detailed education about preterm labor signs and symptoms was tested in a predominantly black, indigent population. One thousand high-risk women were randomized to treatment or control groups. Although more preterm labor was diagnosed and treated in the treatment group, there were no significant differences between the groups with respect to mean birth weight or gestational age, spontaneous preterm delivery rates, or low or very low birth weight rates. The rates of respiratory distress syndrome and fetal and neonatal mortality, although greater in the treatment group, were not statistically different. However, the treatment-group infants had significantly more intracranial hemorrhages and spent more days on ventilators. At this institution, the preterm birth prevention program was not effective. © 1990 The American College of Obstetricians and Gynecologists.