OBJECTIVE The study was conducted because of the still considerable controversies about the variation in obstetric management in breech presentation, in combination with the still increasing caesarean section rate for this type of presentation, as well as doubt concerning the assumed improvement in neonatal and maternal morbidity and mortality obtained by increasing section rate in breeches. METHODS Two hundred and sixty eight consecutive breech presentations in two university hospitals were studied longitudinally and prospectively. In principle, vaginal delivery was attempted in all cases, under expert supervision and electronic monitoring, except in the presence of clear-cut contraindications. X-ray pelvimetry was not used. RESULTS The incidence of caesarean section was 34%. The trial of vaginal delivery succeeded in 64% of the nulliparas and in 60% of the non-frank breeches, without statistically significant differences in outcome measures compared with the multiparous women and the frank group, respectively. Hyperextension of the foetal head rarely occurred. CONCLUSIONS Several classic obstetric factors have a lower discriminating value for the route of delivery than is currently assumed. Progress of labour was an important factor in determining the route of delivery.