The study compares frequencies of birth defects between immigrant groups and the rest of the Norwegian population in Norway and estimates the influence of consanguinity and socioeconomic factors on these frequencies. The authors studied all 1.56 million births in Norway from 1967 to 1993. Of these, 7,494 children had two Pakistani parents, 84,688 had one Norwegian and one immigrant parent, and 25,891 had two immigrant parents from countries other than Pakistan. The risk of birth defects relative to the Norwegian group was 0.98 (95% confidence interval 0.92-1.03) in the group with one foreign and one Norwegian parent, 1.39 (95% confidence interval 1.22-1.60) in the group with two Pakistani parents, and 1.04 (95% confidence interval 0.95-1.14) in the group with two parents from other foreign countries; 0.1% of the Norwegian and 30.1% of the Pakistani children had parents who were first cousins. There was no difference in risk between children of nonconsanguineous Pakistani parents and the other groups. The relative risk of birth defects among children whose parents were first cousins was about 2 in all groups. Among the Pakistani, 28% of all birth defects could be attributed to consanguinity. Low paternal educational level was associated with a slightly increased risk in the Norwegian group, while independent effects of parental educational levels were not found in any other groups.