One of the most important and changing areas of research in paediatric rheumatology is the optimum approach to the treatment of children with chronic arthritis. Until recently all medications for children with arthritis were nonspecific in terms of our understanding, albeit poor, of the pathogenesis of these diseases. Of current therapies, low dose, once-a-week methotrexate has emerged as the therapeutic agent of choice for children who fail to respond adequately to administration of a nonsteroidal anti-inflammatory drug. Thereby, it has displaced the more traditional slower acting anti-rheumatic drugs, although one or more of them are often combined with methotrexate in the polypharmaceutical approach to childhood arthritis. Better and more specific agents are needed, especially for systemic onset disease, unremitting polyarticular involvement, and certain complications such as resistant chronic uveitis. At this time the introduction of the cyclo-oxygenase 2 inhibitors and etanercept (soluble tumour necrosis factoralpha.p75 fusion protein) may herald an era of more specific and effective therapy.