To investigate the therapeutic and economic implications of serum theophylline assay (STA) use in Medicare patients classified in diagnosis-related group 88 as having chronic obstructive pulmonary disease, a retrospective audit was conducted of all serum theophylline assays ordered for these patients in a large teaching hospital during a 6-month period. Based on established pharmacokinetic principles of theophylline therapy, analysis was conducted on rational indication for assay use (n = 146), correct sampling procedure regarding steady-state and peak/trough concentrations (n = 99), and appropriateness of physician response (n = 100). Percentages of appropriate use found for these three parameters were 79.2%, 56.6%, and 85.0%, respectively. Overall compliance regarding all three considerations was 52.1%; this figure declined to 29.9% when admission level assays were excluded from consideration. A need is identified to improve STA sampling procedures in order to insure availability of accurate and valid results on which to make indicated changes in drug therapy.