The class I MHC allele HLA-B27 is highly associated with the human spondyloarthropathies, but the basis for this association remains poorly understood. Transgenic rats with high expression of HLA-B27 develop a multisystem inflammatory disease that includes arthritis and colitis. To investigate whether CD8alphabeta T cells are needed in this disease, we depleted these cells in B27 transgenic rats before the onset of disease by adult thymectomy plus short-term anti-CD8alpha mAb treatment. This treatment induced profound, sustained depletion of CD8alphabeta T cells, but failed to suppress either colitis or arthritis. To address the role of CD8alpha(+)beta(-) cells, we studied four additional groups of B27 transgenic rats treated with: 1) continuous anti-CD8alpha mAb, 2) continuous isotype-matched control mAb, 3) the thymectomy/pulse anti-CD8alpha regimen, or 4) no treatment. Arthritis occurred in approximately 40% of each group, but was most significantly reduced in severity in the anti-CD8alpha-treated group. In addition to CD8alphabeta T cells, two sizeable CD8alpha(+)beta(-) non-T cell populations were also reduced by the anti-CD8alpha treatment: 1) NK cells, and 2) a CD4(+)CD8(+)CD11b/c(+)CD161a(+)CD172a(+) monocyte population that became expanded in diseased B27 transgenic rats. These data indicate that HLA-B27-retricted CD8(+) T cells are unlikely to serve as effector cells in the transgenic rat model of HLA-B27-associated disease, in opposition to a commonly invoked hypothesis concerning the role of B27 in the spondyloarthropathies. The data also suggest that one or more populations of CD8alpha(+)beta(-) non-T cells may play a role in the arthritis that occurs in these rats.