A prospective study in the United Kingdom of 1,634 patients without transplants treated with immunosuppressive drugs (68 percent with azathioprine, 28 percent with cyclophosphamide) found an excess of non-Hodgkin's lymphoma and squamous cell skin cancer, suggesting that the excesses (although larger) of the same malignancies found among transplant recipients are not due solely to the foreign antigens of the graft. A separate analysis of the 643 patients with rheumatoid arthritis found a 13-fold increase of non-Hodgkin's lymphoma (whether treated with azathioprine or cyclophosphamide). This increase is not significantly different from the excess in similarly treated patients with other disorders in the study. In patients with rheumatoid arthritis not receiving immunosuppressive drugs, this excess is greater than that in a Finnish population and lower than that in another United Kingdom population. The findings are consistent with other evidence that immunosuppression favors the development of non-Hodgkin's lymphoma, which includes the excess of malignancies found among transplant recipients, long-term renal dialysis patients, and patients with certain primary immunodeficiency disorders. The higher risk among transplant recipients may reflect the effects of the foreign antigens, the more intensive immunosuppressive therapy, or both of these factors. In addition, the predilection for the brain, which is a well-known feature of the lymphomas after transplantation, may also apply (to a lesser extent) to other patients after immunosuppressive treatment, judging from the increasing numbers of case reports in such patients of this exceedingly rare type of malignancy. In view of the evidence of an increase of non-Hodgkin's lymphoma in rheumatoid arthritis in the absence of immunosuppressive treatment, any additional increase is likely to be small in absolute terms. Nevertheless, it needs to be weighed against the clinical benefits.