In a randomized, controlled trial in five regional centers with state health department clinics, 1458 women at high risk for low birth weight (LBW) outcome received either prenatal interventions provided by nurse-midwives and nurses under their supervision or the standard high-risk prenatal care provided by obstetricians. The intervention administered by the nurse-midwives included patient education to identify the signs and symptoms of preterm labor, activity counseling in response to monitoring of the cervix by frequent examinations, stress reduction by enhancing social support, nutrition counseling with emphasis on weight gain, and substance-abuse counseling. For women in the control group, care was provided by obstetricians according to local standards for the management of high-risk pregnancies. We hypothesized that the LBW rate among live births to women who had received care from nurse-midwives would be lower than that in the control group. Although the LBW rate was lower in the intervention group than in the control group, the observed difference was not statistically significant. Race was not prespecified as a possible effect modifier, but examination of the data post hoc suggested that black women at high statistical risk of giving birth to an LBW infant may have derived benefit from the program. Although the results do not suggest any striking advantage of the nurse-midwifery intervention over standard obstetric care for women at high statistical risk of having an LBW infant, neither do they suggest any disadvantage. Nurse-midwives could provide care to certain populations of high-risk women and facilitate future coverage of these presently underserved populations.