The objectives of treatment in rheumatoid arthritis (RA) are to reduce temporary symptoms due to inflammatory activity and, more importantly, to preserve function. The introduction of potent disease-modifying anti-rheumatic drugs (DMARDs) in recent years has increased the opportunities for effective treatment. However, these treatments come at a substantially higher cost than traditional DMARDs and therefore compete with other essential interventions for limited resources. They have triggered a debate on whether they represent an efficient use of resources, which patients should be treated, when, and for how long. Cost-effectiveness analysis attempts to estimate the trade-offs involved in these decisions and to provide information that can help in making them. However, in chronic progressive diseases, health gains and any potential associated economic benefits are often most evident in the long-term. As a consequence, the impact of new treatments has to be estimated using models that can project available knowledge, such as results from clinical trials or short-term follow-up studies in clinical practice, into the future. These models also allow scenarios to be explored that provide the best value for money, for example by defining subgroups for which treatment is most effective, or criteria that define when treatment should be stopped. Economic evaluation in RA has a long tradition, with the first study performed about 20 years ago. However, with the recent drug introductions, the field has witnessed an explosion of economic studies. Modelling techniques have become more sophisticated to overcome concerns about their validity. At the same time, they may appear less transparent, making it difficult for non-specialists to understand the details. This article, rather than reviewing all published models and comparing them, attempts to illustrate the inputs required for such studies, and the influence that different approaches and datasets can have on the results.