A pilot fetal alcohol syndrome (FAS) surveillance was carried out in four American Indian communities in the Northern Plains by the Aberdeen Area Indian Health Service to determine the incidence of FAS and to evaluate the feasibility of establishing continuing surveillance for FAS. Baseline data on the incidence of FAS would be used by the Indian Health Service to develop and evaluate preventive interventions, including treatment programs for pregnant women who drink alcohol. Four of the 1,022 children included in the project were found to have FAS, a rate of 3.9 per 1,000 live births. The rate is believed to underestimate the true rate of FAS because some low birth weight children were not screened, parents or guardians were reluctant to bring children suspected of FAS for evaluation, clinicians were hesitant to diagnose possible alcohol-damaged children for fear of labeling the child, and some children with FAS died before the diagnosis of FAS could be confirmed. If the rate of FAS is similar for the 39 percent of the infants not screened and for the 25 percent of suspected infants who were not evaluated, a rate of 8.5 cases of FAS per 1,000 live births may be postulated. The authors recommend routine screening of prenatal patients for substance abuse and establishing a tracking system for low birth weight infants suspected to have FAS or other alcohol-related developmental disorders, in an effort to establish more accurate FAS rates. Such a surveillance system would identify women at risk of having alcohol-affected infants so that appropriate treatment and counseling could be provided, possibly reducing the severity of adverse effects of alcohol on their fetuses.