Thiopurine S-methyltransferase (TPMT) catalyses the S-methylation of thiopurine drugs such as 6-mercaptopurine. A common genetic polymorphism for TPMT is associated with large individual variations in thiopurine drug toxicity and therapeutic efficacy. TPMT*3A, the most common variant allele in Caucasians, has two alterations in amino acid sequence, resulting in striking decreases in TPMT protein levels. This phenomenon results, in part, from rapid degradation through a ubiquitin-proteasome-mediated process. We set out to test the hypothesis that chaperone proteins might be involved in targeting TPMT for degradation. As a first step, hsp90, hsp70 and the cochaperone hop were immunoprecipitated from a rabbit reticulocyte lysate (RRL) that included radioactively labelled *3A and wild-type TPMT. TPMT*3A was much more highly associated with all three chaperones than was the wild-type enzyme. The RRL was also used to confirm the accelerated degradation of *3A compared to wild-type TPMT. Treatment of RRL with the hsp90 inhibitor geldanamycin resulted in enhanced association of hsp90 with wild-type TPMT, an observation that correlated with accelerated ubiquitin-dependent degradation of wild-type TPMT. Geldanamycin treatment of COS-1 cells transfected with FLAG-tagged wild-type also resulted in a time and geldanamycin concentration-dependent decrease in TPMT activity and protein, which was compatible with results obtained in the RRL. These observations indicate that TPMT is a client protein for hsp90 and suggest that chaperone proteins, especially hsp90, are involved in targeting both TPMT*3A and, in the presence of geldanamycin, the wild-type allozyme for degradation. Therefore, chaperone proteins play an important mechanistic role in this clinically significant example of pharmacogenetic variation in drug metabolism.