The University Clinic of Navarra is a general teaching hospital with a total of 413 beds, 60 of which are oncology beds. Of the specimens analysed in 1994 by the clinical pharmacokinetics unit, 53.2% were per formed in cancer patients and 27.3% were performed in patients receiving vancomycin. Methods. We established 10 criteria governing both the process and the result in assessing the quality of vancomycin pharmacokinetic monitoring in oncological patients. We performed a retrospective assessment of the data-monitoring follow-up sheets of 100 cancer patients treated with vancomycin and whose serum vancomycin levels were monitored by the clinical pharmacokinetics unit between April 1993 and March 1994. Results. Overall, the degree of compliance with the 10 defined criteria was good, even though 100% compliance was achieved only in criterion 5 (use of the kinetic method) and criterion 8 (repetition of monitoring must be justified). The lowest compliance was obtained in criterion 1 (evaluation of the initial medical prescription) and criterion 6 (the serum creatinine and daily fluid balances, parameters en compassed within the clinical assessment of the pa tient). Conclusions. The detailed analyses of the crite ria performance suggest that the process of serum sample extraction in the wards should be improved and the frequency of serum creatinine testing should be increased. Corrective measures have been pro posed, and once they are developed, accepted, and implemented, the criteria will be reevaluated and a further audit performed to determine the usefulness of the changes made.