Leflunomide has recently been approved by the US Food and Drug Administration for the treatment of rheumatoid arthritis. This approval was based on data from double-blind multicentre trials in the US (US 301; leflunomide versus methotrexate versus placebo) and multicentre European trials (leflunomide versus sulfasalazine versus placebo, and leflunomide versus methotrexate versus placebo). In these trials, leflunomide was superior to placebo and similar to methotrexate or sulfasalazine in efficacy and adverse effects. Both methotrexate and leflunomide retarded the rate of radiological progression, entitling them to qualify as disease-modifying agents (DMARDs). Leflunomide is an immunomodulatory drug that may exert its effects by inhibiting the mitochondrial enzyme dihydro-orotate dehydrogenase (DHO-DH), which plays a key role in the de novo synthesis of the pyrimidine ribonucleotide uridine monophosphate (rUMP). The inhibition of human DHO-DH by A77-1726, the active metabolite of leflunomide, occurs at concentrations (approximately 600 nmol/L) that are achieved during treatment of rheumatoid arthritis. We propose that leflunomide prevents the expansion of activated and autoimmune lymphocytes by interfering with cell cycle progression. This is mediated by inadequate production of rUMP and utilises mechanisms involving the sensor protein p53. The relative lack of toxicity of A77-1726 on nonlymphoid cells may be due to the ability of these cells to fulfil their ribonucleotide requirements by use of the salvage pyrimidine pathway, which makes them less dependent on de novo synthesis.