Bisphenol A (BPA) is used in the manufacture of a range of consumer products, and human biomonitoring studies suggest that exposure to BPA is nearly ubiquitous. We constructed and calibrated a simple pharmacokinetic model to predict urinary concentrations of BPA based on a known initial dose. This descriptive (rather than physiologically based) model has three compartments: "stomach/liver," "blood," and "bladder." We calibrated and validated the model parameters using blood and urine measurements from nine volunteers who consumed 5 mg of d₁₆-BPA. We then applied the model to a second group of eight persons, who supplied full volumes of urine over 7 consecutive days and a diary identifying times and types of food and beverage consumed, to "reconstruct" the time and mass of BPA intakes. These reconstructed daily intakes ranged on average from 60 to 100 ng/kg-day, within the range of, but slightly higher than, those surmised from other studies. About two-thirds of intakes occurred within an hour of reported food or drink consumption, supporting the hypothesis that diet is the main pathway of exposure to BPA. However, one-third of all reconstructed intakes took place outside this time window, suggesting that other sources of BPA exposure may also be relevant.