A randomized controlled trial including 2235 women at high risk of low birthweight was conducted in four Latin American institutions. The objective of this trial was to evaluate a psychosocial support intervention during pregnancy aimed at improving perinatal health and mothers' psychosocial conditions. The core of the intervention was four to six home visits where emotional support, counseling and strengthening of the woman's social network was provided. Outcomes were measured at 36 weeks of pregnancy, post-partum and 40 days after delivery. The intervention was not successful in either altering women's perception of social support and satisfaction with the reproductive experience, as well as maternal and newborn's health care. It is concluded that although high levels of psychosocial distress during pregnancy may play an independent role in determining adverse pregnancy outcomes, this adverse effect does not appear to be ameliorated by psychosocial interventions conducted only during pregnancy, particularly those of a magnitude that can be realistically implemented (in content and frequency) at public care services in most developing countries.