The 'transporter associated with antigen processing' (TAP) gene products are involved in the processing of endogenous peptides that bind to class I molecules. Polymorphism within these genes could alter the level of the immune response, a phenomenon relevant to the development of autoimmune diseases. In this study, we examined the polymorphism of TAP1 and TAP2 genes in patients with rheumatoid arthritis (RA). TAP1 and TAP2 typing was performed for 138 Caucasian RA patients and 100 healthy controls, all originating from eastern France. TAP1 polymorphic residues at positions 333 and 637 and amino acid variants 379, 565, 651 and 665 in the TAP2 gene were found using amplification refractory mutation system-polymerase chain reaction (ARMS-PCR). This method enabled us to determine four TAP1 alleles (TAP1A to TAP1D) and eight TAP2 alleles (TAP2A to TAP2H). All patients and controls had been HLA-DRB1* genotyped. The polymorphic residues TAP1333 and TAP1637 did not show any difference in their distribution between patients and controls. Similar findings were obtained for TAP2379 and TAP2665. However, we found an increased frequency of Thr homozygosity and heterozygosity at position 565 in the TAP2 gene in RA patients (RA vs. controls: 25.3 vs. 14%; P = 0.032; OR = 2.09; CI = 1.01-4.38). Similarly, the prevalence of subjects who were homozygote and heterozygote for Cys651 was increased in the RA group (RA vs. controls: 36.8 vs. 11%; P = 0.02). The dimorphic site TAP2565 defines TAP2D and TAP2E alleles, while the site at position 651 characterizes TAP2F. Thus, we found that TAP2D and TAP2E alleles were more prevalent in RA, but not significantly so (RA vs. controls: TAP2D: 10 vs. 3.6%; P = 0.24; TAP2E: 3.6 vs. 0%; P = 0.19). Similarly, the frequency of TAP2F was higher in RA patients (24.5%) than in controls (11.3%), but this was not significant after correction (P = 0.029; Pcorr = 0.17). Finally, we found no linkage disequilibrium between DRB1* RA-associated alleles and amino acid substitution Thr565 or TAP2D and TAP2E alleles, whereas Cys651 (and TAP2F) was not independent of DRB1*04, a strongly RA-associated allele. Finally, Thr at position 565 in the TAP2 gene was associated with manifestations of disease severity in only a few patients. Examination of TAP1 and TAP2 gene polymorphisms in RA patients revealed an association between a particular amino acid residue, namely Thr565 in the TAP2 gene, and RA. This association was found to be weak and did not seem to be a predictor for the severity of the disease.