A method of bioavailability estimation is presented in which a physiologically based kinetic model of lead kinetics is fit simultaneously to blood and bone lead concentrations after a period of exposure to dietary lead. Optimization of the simultaneous fit, varying only fractional absorption, gives the best estimate of fractional bioavailability for each treatment group. The analysis was applied to data from three separate studies in which rats were fed for 30 consecutive days purified diets containing lead added as lead acetate, mine waste-contaminated test soils, or mine waste itself. Fractional absorption decreased as lead intake increased, regardless of the source of the lead; but the magnitude of this dose dependence was lead source-dependent. There were no differences in lead absorption by male and female rats when lead intake was expressed per unit body weight. Fractional absorption varied from 4 to 5%, at low exposure rates (1–2 mg lead/kg/day) when lead acetate was added to the diet, to 0.24% at a high exposure rate (24 mg/kg/day) when a mine waste-contaminated test soil was added to the diet. Comparison of the results of this analysis with the results of a more conventional analysis, in which the bone and blood lead concentrations were separately compared with bone and blood lead concentrations in rats given daily injections of lead acetate intravenously for 29 consecutive days, demonstrated that the standard analysis failed to reveal the dose dependence of fractional absorption.