Thiopurine drug metabolism is a quintessential case of pharmacogenetics. A wealth of experimental and clinical data on polymorphisms in the thiopurine metabolizing enzyme thiopurine methyl transferase (TPMT) has been generated in the past decade. Pharmacogenetic testing prior to thiopurine treatment is already being practiced to some extent in the clinical context, and it is likely that it will be among the first pharmacogenetic tests applied on a regular basis.We analyzed the published TPMT data and identified some lessons to be learned for the future implementation of pharmacogenetics for thiopurines as well as in other fields. These include the need for comprehensive and unbiased data on allele frequencies relevant to a broad range of populations worldwide. The nature and frequency ofTPMTgene polymorphisms in some ethnic groups is still a matter of speculation, as the vast majority of studies onTPMTallele distribution are limited to only a small subset of alleles and populations. Secondly, an appreciation of the limits of pharmacogenetics is warranted, as pharmacogenetic testing can help in avoiding some, but by far not all adverse effects of drug therapy. An analysis of six clinical studies correlating adverse thiopurine effects andTPMTgenotype revealed that an average of 78% of adverse drug reactions were not associated withTPMTpolymorphisms. Pharmacogenetic testing will thus not eliminate the need for careful clinical monitoring of adverse drug reactions. Finally, a careful approach toward dose increases for patients with high enzyme activity is necessary, as TPMT-mediated methylation of thiopurines generates a possibly hepatotoxic byproduct.