To evaluate the significance of the route of delivery for preterm breech deliveries, a retrospective study was performed on 136 infants in breech presentation, weighing less than 1500 g. Thirty-seven (27%) were delivered vaginally, and the remaining 99 (73%) by cesarean section. Although perinatal mortality was higher in the vaginally delivered infants than in those born by cesarean section (54 versus 37%), statistical significance was not shown. One-fourth of the deaths in the vaginal delivery group occurred intrapartum. Although most previous studies have suggested improved outcome with delivery by cesarean section, the present study fails to demonstrate any statistically significant difference. Whether there is, in fact, no difference in outcome attributable to the route of delivery or a difference exists that is not demonstrable by a retrospective study is an open question. It is concluded that the design of the present retrospective study, as in all published studies on this subject, confounds the results and thereby makes conclusions precarious. The limited value of nonrandomized retrospective studies for the evaluation of a certain obstetric procedure is discussed.