Using patients' genetic and phenotypic profiles to identify the best available drug therapy for individual cases remains a clinical aspiration. Considerable recent research has developed this theme in rheumatoid arthritis (RA). Seven reports have evaluated how the high-risk epitope and other genes influence RA susceptibility and severity. There is strong evidence that the shared epitope influences susceptibility, but inconclusive evidence concerning whether it effects severity. The interleukin-1 gene cluster and IL-4/IL-4 genes influence erosive disease, but their effects are mainly in late RA. A further ten studies examined links with erosive disease. Fewer reports evaluated phenotypes: six studies examined outcome predictions using rheumatoid factor and antikeratin/anticyclic citrullinated peptide antibodies and two looked at conventional clinical measures. Anticyclic citrullinated peptide antibodies are useful in establishing the diagnosis of RA, but rheumatoid factor potentially provides the most useful prognostic information. The value of genotyping RA patients in defining their clinical course and determining the most suitable treatment remains unproven.