OBJECTIVES: Urban Health Care Networks become a strong limitation to guarantee efficient and equitable access, especially in developing countries.  Lack of formal protocols in referral procedures, insufficient coordination among levels of care and limited coordination in human resources across health care system allows duplication of clinical studies, informality in the decision making process within the network, and higher-than-expected health risks, affecting existing financial protection mechanisms. In particular, perinatal healthcare networks involve the analysis not only of clinical performance in hospitals and health care centers, but also the ability to improve mothers’ behavior during pregnancy, by applying safe preventive care procedures. This study analyzes two mayor public hospitals in the Province of Buenos Aires, Argentina, searching for quality of referral procedures METHODS: The descriptive and logistic analysis identifies three sets of variables: mother’s background and risk factors, pregnancy characteristics and supply-side infrastructure. Using an original database on birth attendance at the hospital head-of-network (Perinatal Information System, designed by the World Health Organization) of circa 16,500 births, as well as structured questionnaires distributed among human resources in both hospitals, the study allows to identify hospital ability to manage network risks, failures arising from the health care system and join responsibilities between formal healthcare system and the family. RESULTS: The study allows the definition of two sets of observations based on number of medical controls during pregnancy, offering the chance of additional comparisons of explanatory variables by including test of differences in mean values. CONCLUSIONS: Results show the presence of multiple failures in urban health care networks, with specific risk factors teenage pregnancy and distance to formal health services affecting hospitalization rates, newborn risks and counter-reference behavior.