Objective To develop an indicator of maternal near miss as a proxy for maternal death and to study its association with maternal factors and perinatal outcomes. Methods In a multicenter cross-sectional study, we collected maternal and perinatal data from the hospital records of a sample of women admitted for delivery over a period of two to three months in 120 hospitals located in eight Latin American countries. We followed a stratified multistage cluster random design. We assessed the intra-hospital occurrence of severe maternal morbidity and the latter's association with maternal characteristics and perinatal outcomes. Findings Of the 97 095 women studied, 2964 (34 per 1000) were at higher risk of dying in association with one or more of the following: being admitted to the intensive care unit (ICU), undergoing a hysterectomy, receiving a blood transfusion, suffering a cardiac or renal complication, or having eclampsia. Being older than 35 years, not having a partner, being a primipara or para > 3, and having had a Caesarean section in the previous pregnancy were factors independently associated with the occurrence of severe maternal morbidity. They were also positively associated with an increased occurrence of low and very low birth weight, stillbirth, early neonatal death, admission to the neonatal ICU, a prolonged maternal postpartum hospital stay and Caesarean section. Conclusion Women who survive the serious conditions described could be pragmatically considered cases of maternal near miss. Interventions to reduce maternal and perinatal mortality should target women in these high-risk categories.Une traduction en français de ce résumé figure à la fin de l'article. Al final del artículo se facilita una traducción al español. املقالة. لهذه الكامل النص نهاية يف الخالصة لهذه العربية الرتجمة