Established loci for rheumatoid arthritis (RA), including HLA-DRB1 and PTPN22, do not fully account for the genetic component of susceptibility to the disease. One possible source of as yet undiscovered susceptibility genes are those mediated through effects of rare variants. We present a novel method for gene-based genome-wide scans of whole-genome association (WGA) data to identify accumulations of rare variants associated with disease. We apply our method to WGA SNP genotype data obtained from 868 RA cases and 1194 controls. Our results highlight novel putative RA susceptibility genes that have not previously been identified in large-scale WGA studies.