Published data concerning proposed antenatal Doppler criteria for intrauterine growth retardation (IUGR) were critically analyzed using the same technique that has been previously applied to conventional (non-Doppler) sonographic criteria for IUGR. Fifteen studies of Doppler criteria, each conducted in such a way that sensitivity and specificity could be determined, were identified through a literature review. These studies encompassed a variety of arterial waveform and volume blood flow criteria. Bayes' theorem was then used to determine the positive predictive value of each criterion if used as a screening test for IUGR. Criteria involving vessels that have been studied by at least two groups, including those involving umbilical artery waveform, uterine arcuate artery waveform, and umbilical vein volume blood flow, had positive predictive values that fell in a low range (17-57%) similar to that previously found for conventional criteria. Two criteria, using fetal internal carotid artery and thoracic aortic waveforms, had higher positive predictive values (66 and 100%, respectively) in initial studies, but the data have yet to be replicated by other groups. It is concluded that no Doppler criterion has yet been established as a clinically useful method to screen for IUGR antenatally.