The developmental toxicity potential of a scrubbing solution used extensively in petroleum refineries to remove CO2 from hydrogen gas streams was evaluated via inhalation. Pregnant female CD (Sprague-Dawley) rats were exposed to aerosols of a "used" scrubbing solution at 0.05, 0.1, 0.2, or 0.3 mg/l for 6 h/d on d 6-19 of pregnancy. Control animals were exposed to filtered air under the same exposure conditions. Dams were sacrificed on d 20 of pregnancy and a laparohysterectomy was performed. The mass median aerodynamic diameter of the aerosol revealed that all particles ranged from 1.6 to 2.8 mum, with geometric standard deviations between 2.0 and 2.3 mum. The overall pregnancy rate was high (>95% ) and equivalent across all groups. All pregnant dams had live litters, and 22-24 litters were examined in each group. Treatment-related clinical signs consisted of rales, observed at all exposure levels, and gasping noted only at the 0.3 mg/l exposure level. The occurrence of rales was presumably a localized effect on the respiratory tract and likely due to the irritating properties of the scrubbing solution. Maternal toxicity was exhibited in the 0.3 mg/l group, including reduced body weight, weight gain, and food consumption and one possible treatment-related death on gestation d 17. At scheduled necropsy, there were no treatment-related gross pathological observations and no statistically significant reproductive and developmental effects. The incidences of fetuses with skeletal variations involving the sternum were clustered in two litters at the highest exposure level with atypically low term fetal body weights. Under the conditions of this investigation, potassium carbonate scrubbing solution is not a selective developmental toxicant.