This study was conducted to determine the extent of arsenic (As) absorption from soil and house dust impacted by smelter activities near Anaconda, Montana. Female cynomolgus monkeys were given a single oral administration via gelatin capsules of soil (0.62 mg As/kg body wt) or house dust (0.26 mg As/kg body wt), or soluble sodium arsenate by the gavage or intravenous route of administration (0.62 mg As/kg body wt) in a crossover design with a minimum washout period of 14 days. Urine, feces, and cage rinse were collected at 24-hr intervals for 168 hr. Blood was collected at specified time points and area under the curves (AUCs) was determined. Arsenic concentrations for the first 120 hr, representing elimination of greater than 94% of the total administered dose for the three oral treatment groups, were < 0.021 to 4.68 micrograms/ml for the urine and < 0.24 to 31.1 micrograms/g for the feces. In general, peak concentrations of As in the urine and feces were obtained during the collection intervals of 0-24 and 24-72 hr, respectively. The main pathway for excretion of As for the intravenous and gavage groups was in the urine, whereas for the soil and dust groups, it was in the feces. Mean absolute percentage bioavailability values based on urinary excretion data were 68, 19, and 14% for the gavage, house dust, and soil treatments, respectively, after normalization of the intravenous As recovery data to 100%. Corresponding absolute bioavailability values based on blood were 91, 10, and 11%. The bioavailability of soil and house dust As relative to soluble As (by gavage) was between 10 and 30%, depending upon whether urinary or blood values were used. These findings suggest that risks associated with the ingestion of As in soil or dust will be reduced compared to ingestion of comparable quantities of As in drinking water.