A study of the admissions to the University of Bergen, Norway, hospital for ectopic pregnancy during the 1965-78 period was conducted. A retrospective study of patients' records for cases occurring during 1965-71 was made. For the later period, the study was made prospectively by registering every new patient and emphasizing analysis of etiological and clinical factors. Of the 450 ectopic pregnancies occurring during this period, 105 had an IUD in place. The incidence of ectopic pregnancies in women with IUDs increased in the last 5 years. Lippes loop was the IUD most often involved early in the study and the Copper T and Copper 7 were more prevalent later on, reflecting device popularity. The IUD group had significantly fewer etiological findings for ectopic pregnancy than the non-IUD group. In other words, there were many less incidences of pelvic inflammatory disease, other pelvic adhesions, and earlier spontaneous abortions in the IUD group than in the non-IUD group. Patients with an IUD in place tended to register symptoms earlier in the ectopic pregnancy but to be in worse condition at the time of admission to the hospital. Localization of the ectopic pregnancy is tabulated. There was a statistically significant greater incidence of ovarian pregnancies among the IUD women, which is possibly explained by the fact that the antifertility effect of the IUD is strongest in the endometrial cavity, weaker in the tube, and absent beyond the tube.