OBJECTIVE To determine the relationship of what has been called pre-Columbian Old World rheumatoid arthritis (RA) to the RA identified in pre-Columbian North America. METHODS All published claims of pre-Columbian Old World RA were reviewed against the established North American standard for its recognition in archeologic sites. Those characteristics included polyarticular symmetrical marginal erosions [in the absence of subchondral erosions, peripheral joint fusion, or axial skeletal involvement (C1-2 excepted)], but requiring the presence of perilesional osteopenia on radiographic examination. T test and Fisher's exact test were used to assess the significance of the extent of joint distribution and the presence of subchondral erosions, peripheral joint fusion, and axial disease in the Old World cases that some have claimed represent RA. RESULTS Old World reports of alleged RA often describe isolated bones or isolated "finds" without epidemiologic consideration. Subchondral erosions were present in 95%. The 2 cases without subchondral erosions had peripheral joint fusion and axial joint disease. Peripheral joint fusion and axial joint involvement were present in almost all cases. Perilesional sclerosis was actually quite prominent, as was other evidence of reactive new bone formation, but not perilesional osteopenia. CONCLUSION As the pre-Columbian Old World erosive arthritis is clearly a different phenomenon from what has been documented in the New World, the issue appears to relate to criteria for naming RA. There clearly are 2 distinct groups that some classify under the broad banner of RA. As the Old World variety is indistinguishable from spondyloarthropathy, it is suggested that the Old World cases should be recategorized with spondyloarthropathy and that only the variety reported in archeologic sites in North America be classified as RA.