Abstract: The developmental toxicity of arsenic is not as well characterized as other metals such as lead or mercury. Many previous animal studies have used an acute exposure paradigm, which does not model chronic, low-level human exposure. The following study administered 10, 20, 40, 80 or 100 ppm sodium arsenite in drinking water to pregnant C57BL6/J mice. Adipose, blood, brain, breastmilk in stomach, kidney and liver tissues were collected from male and female offspring on postnatal day (PND) 1 and 21 to allow for disposition comparisons between tissues, sexes and across time. The 100 ppm dose was foetotoxic. Significantly fewer female pups were born in litters exposed to 80 ppm, while significantly more male pups were born in litters exposed to 20 ppm. Total arsenic levels differed between tissues with the highest levels in the brain and kidney in PND1 offspring. Levels were higher on PND1 than PND21, and there were few sex differences. The dose–response relationships in PND1 tissues were curvilinear, but in PND21 liver and kidney tissues, arsenic levels in control animals were significantly higher than levels in exposed animals. The tissue and age-specific disposition suggests that common biomarkers such as blood and urinary arsenic are not accurate predictors of levels in sensitive organs such as the brain.