This review of eight randomized controlled trials assessed two different types of home visits during pregnancy: (1) those offering social support to high-risk women; and (2) those providing medical care to women with complications. In both categories, pregnancy outcome was not improved when women received home visits. The summary odds ratio for preterm delivery (< 37 weeks) was 1.0 (95% CI: 0.8 to 1.1). Nor did the home visits decrease the rate of hospital admission for women with complications (mainly threatened preterm labor or toxemia); the corresponding summary odds ratio was 0.9 (95% CI: 0.7 to 1.2). Nevertheless in some trials home visits had positive effects on women (medical knowledge, support levels, health habits, and satisfaction). The randomized controlled trials provide little evidence that programs offering home visits are effective in improving either pregnancy outcome or the use of health services. A better integration of hospital and home services might allow a more rational use of health services for women with complications. In addition, we need to define more precisely the content of home visits providing social support. For this, further research is required on how emotional support, health education, and advice influence the health of women and infants and mother-child interactions.