Timed-pregnant CD rats and CD-1 mice were dosed by gavage with ethylene glycol (EG) in distilled water on gestational days (gd) 6 through 15 (0, 1250, 2500, or 5000 mg kg-1 day-1 for rats; and 0, 750, 1500, or 3000 mg kg-1 day-1 for mice). Females were observed daily during treatment, but no maternal deaths or distinctive clinical signs were noted. Dose-related decreases in maternal weight gain during treatment were significant at all doses in rats and at the mid and high doses in mice. Gravid uterine weight was reduced in both species at the mid and high doses, and corrected maternal gestational weight gain showed a significant decreasing trend. At termination (gd 20, rats; gd 17, mice), the status of uterine implantation sites was recorded, and live fetuses were weighed and examined for external, visceral, and skeletal malformations. Dose-related increases in postimplantation loss per litter were observed in both species with the high dose significantly above controls only in rats. Fetal body weight per litter was significantly reduced at the mid and high doses in rats and at all doses in mice. The percentage of malformed live fetuses per litter and/or the percentage of litters with malformed fetuses was significantly elevated in all EG dose groups and greater than 95% of litters were affected at the high dose for each species. A wide variety of malformations were observed; the most common in both species were craniofacial and neural tube closure defects and axial skeletal dysplasia. EG produced severe developmental toxicity in two rodent species at doses that apparently failed to produce any serious maternal effects.