In 1983 a protocol was established at our institution that used computed tomographic pelvimetry to evaluate patients presenting with a singleton term frank breech fetus for an attempt at vaginal delivery. The criteria for a trial of labor were singleton fetus, term gestation (37 to 42 weeks' gestation from the first day of the last menstrual period), frank breech presentation, estimated fetal weight 2000 to 4000 gm, non-extended fetal head, and adequate computed tomographic pelvimetry. Two digital radiographs and a tomographic cross-section were taken of each patient, i.e., an anteroposterior view, a lateral view, and an axial section through the femoral head at the level of the fovea capitalis. Adequate pelvimetry was defined as: anteroposterior diameter of the inlet greater than or equal to 10.0 cm, transverse diameter of the inlet greater than or equal to 11.5 cm, transverse (interspinous) diameter of the midpelvis greater than or equal to 9.5 cm, and posterior sagittal diameter of the midpelvis greater than or equal to 4.0 cm. The purpose of this prospective analysis was to determine the obstetric and perinatal outcome of those breech fetuses that were evaluated with these criteria during the study period of January 1984 through July 1989. During this period the incidence of breech deliveries at our institution was 2.71% (394/14,542). Of these 394 breech deliveries, 122 patients underwent computed tomographic pelvimetry. Eighty-five patients had adequate pelvimetry, fulfilling the protocol criteria, and formed the study group. Eighty-one percent (69/85) of the study group had successful vaginal deliveries (group 1). Nineteen percent (16/85) required cesarean delivery after a trial of labor (group 2) (10 with fetal distress and six with arrested labor disorders). Of the 37 patients who had inadequate computed tomographic pelvimetry and underwent cesarean delivery without a trial of labor (group 3), 54% (20/37) had an extended fetal head, 21.6% (8/37) had an inadequate transverse diameter of the inlet, 13.5% (5/37) had an inadequate midpelvic posterior sagittal measurement, and 10.8% (4/37) had an inadequate interspinous diameter. Perinatal outcome including Apgar scores, cord gases, length of hospital stay, neonatal complications, was evaluated. No difference in infant complications was noted between the group delivered vaginally and the groups delivered abdominally. The only maternal complications in our study group were chorioamnionitis, endomyometritis, and postpartum anemia. There were no significant differences in the incidence of chorioamnionitis among the three groups of patients.(ABSTRACT TRUNCATED AT 400 WORDS)