Perinatal health problems are a public health priority in Latin America. Among the identified risk factors, psychological and social conditions play a crucial role. However, care during pregnancy and delivery in the region is usually hospital-centered and does not address women's psychological and social conditions. The preeminence of research on perinatal health, along with the necessity for testing interventions that represent alternative models to improve women's health, gave the Latin American Network for Perinatal and Reproductive Research grounds to develop a multicenter randomized controlled trial to evaluate a program of social support and health education during pregnancy. The conceptual framework for this study was based on an ecological model of social support, i.e. a model in which social support and health education play a synergistic role and are meant to modify stressful situations and negative health-related behaviors. The target population consisted of women attending obstetric hospitals before the 22nd gestational week, at high psychological and social risk (n = 2236). The intervention consisted of four to six home visits, carried out by social workers, and had four main components: the reinforcement of pregnant women's social support network, emotional support, health education, and the improvement of health services utilization. The main foci of the intervention were determined after an ethnographic study was carried out to identify stress-producing situations and needs for support during pregnancy. Besides the home visits, the program also offered a hot-line, an office in the hospital, a specially designed poster and booklet, and a 'guided tour' of the health institution. Since this was a multicenter trial, the program's standardization was a crucial methodological aspect that was achieved through the training course for the home-visitors team. Biological and psychosocial outcomes were measured in both experimental and control groups at the 36th week of gestational age, post-partum and at the 40th day after delivery. The attributes of the multicenter population showed an important variability, reflecting differences in the countries or hospitals' population prevalent attributes. The results of the program's implementation were analyzed, demonstrating that home visitors adapted topics discussed during the interviews to the women's conditions and the stage of pregnancy during which the visit took place.(ABSTRACT TRUNCATED AT 400 WORDS)