The NIH has recently reported results of an NCI study demonstrating an increase in incidence of liver, lung, and genitourinary tumors in the offspring of mice receiving high doses of AZT during the last trimester of gestation. A second study from Glaxo Wellcome, also in mice and designed to more closely simulate doses (one-twelfth to one-fiftieth daily doses in NCI study) and schedules utilized in actual clinical practice, found no increase in tumors in offspring exposed to AZT. No tumors have been identified to date in studies of approximately 1,000 children exposed to AZT and followed for an average of 3 years. An independent panel of basic and clinical researchers, epidemiologists, a bioethicist, and HIV-infected women was convened by the NIH on January 14, 1997. This panel noted that there are major differences between humans and mice in terms of maternal-fetal relationships and the pharmacokinetics of AZT making the implications of these findings unclear. Although this new information warrants attention and careful follow-up, the consensus at this time is that current public policy recommendations and treatment guidelines should remain in place. The significant clinical benefits in reduction of perinatal transmission with AZT given during pregnancy far outweighs the still hypothetical risks suggested by this study. It is important to include this information about possible carcinogenesis in discussions with all HIV-infected pregnant women in the course of counseling about treatment interventions, whether in clinical practice or clinical trials.