Currently the diagnosis of rheumatoid arthritis (RA) may be difficult; the ACR criteria appear most sensitive and specific in long-standing disease. Without clear definition or diagnostic criteria for early disease it is difficult to define late or established RA. The distinction between early and established RA has been further blurred by recent imaging studies that suggest even in what is currently termed early disease, there is evidence of joint damage. The natural history of RA suggests that most patients with clinic-diagnosed RA have a progressively disabling course, but evidence is growing that modern therapeutic strategies result in better long-term outcomes, especially when applied early in the disease course. In established disease, quantitative markers such as C-reactive protein (CRP) give prognostic information, but in the pre-erosive, early phase of the disease the qualitative markers such as rheumatoid factor (RF) and shared epitope are crucial. As rheumatologists, our major aims must remain: (1) to diagnose the disease as early as possible; (2) to identify those patients with poor prognosis who will benefit most from targeted therapy; and (3) to aim for more intensive disease control irrespective of disease duration.