OBJECTIVE The relative effectiveness of bilateral and unilateral gamete intrafallopian transfer (GIFT) was compared. Bilateral GIFT, where possible, was the preferred method of transfer and occurred in 328 patients. One hundred sixty-seven patients undergoing unilateral GIFT were placed into one of seven categories based on the indication for unilateral GIFT. Three categories included patients with either historical or laparoscopic evidence of unilateral tubal abnormalities, two categories included patients without tubal disease, and two categories contained patients in whom the physician could not choose between unilateral or bilateral GIFT. RESULTS Results demonstrate that unilateral GIFT in patients with either historical or laparoscopic evidence of unilateral tubal disease had significantly lower pregnancy and implantation rates than patients undergoing bilateral GIFT or those undergoing unilateral GIFT without evidence of tubal disease (P < 0.01). CONCLUSIONS The presence of unilateral tubal abnormalities predicts a lower likelihood of pregnancy in patients undergoing unilateral GIFT through the tube presumed to be normal.