Diethylene glycol (DEG; CAS Number 111-46-6) is a widely used industrial liquid chemical with a potential for human exposure. In view of the established teratogenic effects caused by ethylene glycol in laboratory animals, the developmental toxicity of DEG was investigated in mice and rats, species known to be sensitive to the developmental toxicity of ethylene glycol. Timed-pregnant CD-1 mice and CD rats were dosed daily by gavage with undiluted DEG over gestational days (gd) 6-15. Based on probe studies, mouse dosages were 0 (distilled water), 559, 2795 and 11,180 mg/kg/day, and those for rats 0, 1118, 4472 and 8944 mg/kg/day. They were examined daily for clinical signs of toxicity, and body weights, food consumption and water consumption measured periodically throughout gestation. At necropsy, on gd 18 (mice) or gd 21 (rats), dams were examined for gross pathology and body, gravid uterus, liver and kidney weights were measured. Maternal rat kidneys were examined histologically. Fetuses were weighed, sex determined, and examined for external, visceral and skeletal variations and malformations. With mice there was maternal toxicity at 11,180 mg/kg/day (mortality, signs, increased water consumption) and at 2795 mg/kg/day (increased water consumption). Implantations were comparable across all groups. Fetal body weights were significantly reduced at 11,180 mg/kg/day. There were no increases in variations or malformations, either total, by category, or individually. With rats, maternal toxicity was present at 8944 mg/kg/day (mortality, signs, reduced body weight gain, reduced food consumption, increased water consumption, increased liver weight, increased kidney weight, and renal histopathology), and 4472 mg/kg/day (increased water consumption). There were no treatment-related effects on corpora lutea or implantations. Fetal body weights were reduced at 8944 mg/kg/day. There were no significant effects with respect to total or individual external or visceral variations. Individual skeletal variations were significantly increased at 8944 mg/kg/day (poorly ossified interparietal, poorly ossified thoracic centra number 10 and number 13, and bilobed thoracic centrum number 10) and 4472 mg/kg/day (split anterior arch of atlas and bilobed thoracic centrum number 10). This pattern of delayed ossification is consistent with reduced fetal body weight. Malformations, total, by category, or individually, were similar between the control and DEG groups. Thus, under the conditions of these studies, the no-observed-effect-level (NOEL) for DEG given by gavage over gd 6-15 was 559 mg/kg/day with the mouse and 1118 mg/kg/day with the rat for maternal toxicity, and 2795 mg/kg/day with mice and 1118 mg/kg/day with rats for developmental toxicity (fetotoxicity). There were no indications of embryotoxicity or teratogenic effects at any dosage in either species.