A wide range of interindividual variability of 5-fluorouracil (5-FU) pharmacokinetics exists after bolus administration. The degree to which this variability in 5-FU exposure impacts upon the response and toxicity of the drug has not been determined. The area under the concentration time curve (AUC) is a commonly used indicator of exposure, but normally requires the collection of 8-10 timed blood samples after i.v. bolus administration. This presents difficulties if large-scale population samplings are required. This study involved the development and testing of a strategy to estimate AUC from a limited number of blood samples in patients with gastrointestinal and breast cancer. The optimal single time point for AUC estimation was 0.17 h (r2 = 0.954). Addition of the 0.75 h time point significantly improved predictability (r2 = 0.983). Addition of a third or fourth time point did not provide further benefit. These models were then tested separately in a group of women who received a higher dose of 5-FU. The two data points model performed significantly better than the single time point model (r2 = 0.70 and 0.85, respectively). The AUC of standard dose 5-FU after bolus administration can be reliably estimated from two timed samples taken 10 and 45 min after injection.