Allopurinol is converted almost completely into a single active metabolite, oxipurinol, which has the same therapeutic pattern but a much longer elimination half-life than the parent compound. Therefore both allopurinol and oxipurinol were evaluated in our bioequivalence study in healthy volunteers comparing two allopurinol brands. Bioequivalence determination was based on the 90% confidence intervals (CI) of the area under the plasma concentration time curve from time zero to infinity (AUC0-infinity), of the area from time zero to the last measurable plasma concentration (AUC0-t (last)), and Cmax. Because of the lack of compound-specific criteria we used conventional limits for the bioequivalence range. Under these conditions the brand chosen as test preparation was judged to be bioequivalent to the reference form with respect to the extent of bioavailability, AUC0-infinity, and AUC0-t (last) of the parent drug. The CI of Cmax of allopurinol slightly exceeded the upper limit of 130%, so that bioequivalence was not confirmed with regard to the rate of bioavailability of the parent compound. The CI values of both AUC and Cmax of the active metabolite were tighter than those of allopurinol. In addition, the CI values of Cmax of oxipurinol were smaller than those of the corresponding AUC. As a consequence the test drug can clearly be accepted as bioequivalent, based on metabolite data. Since the active metabolite is of greater therapeutic significance than the parent drug, assessment of the bioequivalence of allopurinol preparations needs to be based on oxipurinol rather than allopurinol.(ABSTRACT TRUNCATED AT 250 WORDS)