Aim: External cephalic version (ECV) is an effective and safe technique for avoiding breech presentation at birth. However, it continues rejected by many women. The aim of this study is to develop a predictive model of success of external cephalic version, determine the safety of the technique and perinatal outcomes after successful version. Methods: Data from 317 versions performed over a 6-year period were collected. Different clinical and ultrasound variables, complications, vaginal delivery after successful version and perinatal outcomes were analyzed. Results: The overall success rate was 72% (229 of 317 versions). The variables most related to success were parity, placental location, amniotic fluid volume, fetal sex, fetal head palpation and descent of the presenting part. A model for calculating the probability of success was developed in which to input parity, placentation and amniotic fluid data. The model correctly classified 98.8% of successful technique and 74% of all women. Complications were very few and mostly mild. Of women who had success, 77% (163 of 212) had a vaginal birth. No differences between neonatal outcomes were found. Conclusion: External cephalic version is a successful, safe technique with a high rate of subsequent vaginal delivery. A success prediction model based on some very easily obtained variables can personalize the probability of success.