OBJECTIVE: We sought to investigate what aspects of the stillbirth evaluation are considered to be essential and what tests can potentially be eliminated. STUDY DESIGN: A retrospective analysis of 745 stillbirths occurring from January 1990 to December 1994 was conducted. A stillbirth was defined by an estimated gestational age >20 weeks' gestational age or fetal weight >500 gm. We attempted to arrive at an apparent cause for each stillbirth after evaluation of genetic or chromosomal abnormalities, obstetric history, maternal medical illnesses, laboratory tests, autopsy findings, and placental pathologic conditions. RESULTS: We found that the most important aspects of stillbirth evaluation were placental pathologic conditions and autopsy. When the placenta was examined, a significant abnormality was detected in 30% (160 of 529) of the cases. When autopsy was performed, only 31% of fetal deaths (142 of 462) were unexplained; however, when no autopsy was performed, 44% (125 of 283) were unexplained (p = 0.0002). The following laboratory evaluations that were routinely performed were found to yield little definitive information: antinuclear antibody testing, Kleihauer-Betke test, and screening for congenital infections (toxoplasmosis, other viruses, rubella, cytomegalovirus, and herpes simplex virus). Overall, 36% (267 of 745) of stillbirths still remained unexplained despite a thorough evaluation in most cases. CONCLUSION: The causes of stillbirth are many and varied, with a large proportion having no obvious cause. As this study demonstrates, certain laboratory tests can be eliminated in the workup of fetal death. In the evaluation of stillbirth a complete systematic method that incorporates placental pathologic conditions, as well as autopsy findings, should prove to be beneficial. (Am J Obstet Gynecol 1998;178:1121-5.)