In the past few years there has been considerable debate regarding the optimal therapy for patients with RA. The arguments have been phrased in a variety of ways and frequently involve the concept of therapy with multiple drugs. The underlying assumption is that current therapy is inadequate and that the whole therapeutic approach to RA needs rethinking. The evidence on which these assumptions are made is reviewed here. Additionally, this paper argues for a more discriminating approach using prognostic indicators to identify the minority of patients with persistent unresponsive disease (and poor outcome) for whom more aggressive therapy should be reserved.