Although stillbirth is a common event, few programs have previously been established for the comprehensive etiologic investigation of intrauterine death. Fewer still have been prospective, unbiased in ascertainment, and consistent in protocol utilization. The Wisconsin Stillbirth Service Program was established in 1983 as a unique model for the investigation of the causes of stillbirth. This community-based, University-supported model for health care delivery is here described. Through it more than a thousand infants have been etiologically investigated. We demonstrate that a community-based program of stillbirth assessment can succeed, that compliance with recommended protocols is high and that a specific primary cause of fetal death can be demonstrated in about 40% of referrals. A majority of the established causes of intrauterine death are fetal etiologies. Furthermore, it appears that there are no substantial referral biases. Lack of such biases, together with the prospective, protocol driven nature of the program provides a unique population upon which to base estimates of the frequency of various etiologic diagnoses and classes and categories of cause.