High perinatal mortality has impaired the reproductive performance of female squirrel monkeys in captivity. A pelvimetry method as a potential predictor of pregnancy outcome was devised to evaluate the bony pelvic structures of females for which records of previous pregnancies were available. By utilizing lateral and anteroposterior radiographic views of the pelvis, the inlet, the midpelvis, and the outlet were measured. In the colony, 21 females had livebirths, 13 delivered dead term fetuses (stillbirths), and six aborted. Comparison of the pelvic outlet of the females with liveborn fetuses (1.81 +/- 0.12 cm) with those which had stillbirths (1.64 +/- 0.09 cm) revealed a highly significant difference (p less than 0.001). Discriminant analysis demonstrated a 95.2% predictability of females which had liveborn and a 92.1% rate in females with stillborn fetuses. Pelvimetry may be a useful tool in predicting pregnancy outcome, as narrow pelvic outlets were consistently observed in females that delivered stillborn infants.