OBJECTIVE To determine whether luteal phase defect (LPD) may be an etiologic factor in ectopic pregnancy (EP). METHODS All patients who were seen over a 6-year period with the chief complaint of infertility underwent an extensive infertility workup and were followed prospectively. The diagnoses of the causes of infertility were assigned retrospectively. METHODS Two hospital-based tertiary care reproductive endocrine-infertility units. METHODS A total of 1,077 infertility patients were evaluated. Of the 633 who became pregnant, the infertility had been due to LPD in 51 and to anovulation in 210. METHODS All the infertility patients who became pregnant were followed to determine whether they miscarried, developed an EP, or had a viable birth. The incidence of EP and miscarriage in the patients whose infertility was found to be due to LPD were compared with a control group in whom the infertility was due to anovulation. RESULTS The EP rate in the patients with LPD was significantly higher than in a control group whose infertility was due to anovulation (6 of 51 pregnancies versus 6 of 210 pregnancies, respectively). The spontaneous abortion rate in LPD cases also was highly significantly greater than in the control group (19 of 51 pregnancies versus 12 of 210 pregnancies, respectively). The EP and spontaneous abortion rates also were higher in patients with LPD who were untreated than in those who were treated. CONCLUSIONS This study suggests that there is a significantly increased incidence of tubal EP in patients with LPD and that when patients with LPD become pregnant early ultrasound should be performed to rule out EP. The study also indicated that spontaneous abortion occurs in a significantly high percentage of LPD cases.