In humans, early life exposure to inorganic arsenic is associated with adverse health effects. Inorganic arsenic in utero or in early postnatal life also produces adverse health effects in offspring of pregnant mice that consumed drinking water containing low part per billion levels of inorganic arsenic. Because aggregate exposure of pregnant mice to inorganic arsenic from both drinking water and food has not been fully evaluated in experimental studies, quantifying arsenic exposure of the developing mouse is problematic. Here, we determined levels of total arsenic and arsenic species in natural ingredient rodent diets that are composed of many plant and animal-derived foodstuffs and in a purified ingredient rodent diet that is composed of a more restricted mixture of foodstuffs. In natural ingredient diets, total arsenic levels ranged from ∼60 to ∼400 parts per billion, and in the purified ingredient diet, total arsenic level was 13 parts per billion. Inorganic arsenic was the predominant arsenic species in trifluoroacetic acid extracts of each diet. Various exposure scenarios were evaluated using information on inorganic arsenic levels in diet and drinking water and on daily food and water consumption of pregnant mice. In a scenario in which pregnant mice consumed drinking water with 10 parts per billion of inorganic arsenic and a natural ingredient diet containing 89 parts per billion of inorganic arsenic, drinking water contributed only ∼20% of inorganic arsenic intake. Quantitation of arsenic species in diets used in studies in which drinking water is the nominal source of arsenic exposure provides more accurate dosimetry and improves understanding of dose-response relations. Use of purified ingredient diets will minimize the discrepancy between the target dosage level and the actual dosage level attained in utero exposure studies designed to evaluate effects of low level exposure to inorganic arsenic.