in the study of fetal deaths. For a variety of reasons adequate records are difficult to obtain. Complete medical histories do not exist for most of the population, and the official records available are usually of doubtful accuracy. Reconstructing pregnancy histories has not proved a satisfactory substitute. Early miscarriages are difficult to identify, and problems of memory and willingness to report fetal deaths are large. More important, among fetal deaths which are reported, induced abortions are difficult to separate from spontaneous abortions.