Determinations of single serum drug concentrations are useful in monitoring drug therapy. A mean serum level would supply more information but is expensive and laborious because of the multiple blood samples and assays, calculation of area under the curve (AUC) by the trapezoidal rule, and division of the AUC by the time interval during which the samples were drawn. A method was devised that pools aliquots from individual serum samples taken during the testing period to form on composite sample. A single assay of this sample provides the mean serum level of the testing period. The method was successfully tested with amaranth and then applied to valproic acid serum levels. AUC and mean serum levels were determined by the standard procedure of assays of multiple samples and the trapezoidal rule. Mean serum level was also determined by the pooled sample technique. The correlation coefficient for the comparison of the mean serum levels determined by the two techniques is 0.907 (p less than 0.001). There was no difference between the estimates of the mean serum levels by Student's paired t test (t = 0.693, p greater than 0.2). The good correlation and lack of difference between the conventional method and the pooled sample method indicates that the method is valid.