OBJECTIVES: We developed a cost-effectiveness model for inflammatory arthritis (IA) patients who were suspected of having rheumatoid arthritis (RA), and analysed the simulated costs and effectiveness of different diagnostic test strategies. METHODS: A decision tree to classify patients as true positive, false positive, true negative, and false negative, followed by a patient-level state transition model was developed. Disease progression was modelled as changes in disease activity (DAS28) over time, which were linked to costs and health-related quality of life obtained from a usual care IA cohort and an early RA trial. We modelled that an earlier diagnosis would improve early treatment, which in turn would reduce DAS28 by 0.2, thereby postponing the start of biologic treatment. Three new diagnostic test strategies (add-on for all patients, add-on for intermediate-risks, and replacement) evaluating four different tests (B-cell, MRI, IL-6 serum levels and genetic assay) were compared with the ACR/EULAR 2010 RA classification criteria representing the current test strategy. The time horizon was 5 years. Probabilistic sensitivity analyses, headroom analyses, and exploratory univariate sensitivity analyses were performed. RESULTS: A B-cell test (sensitivity: 0.60, specificity: 0.90, costs: €150) was the most cost-effective test in all three strategies. MRI (sensitivity: 0.90, specificity: 0.60, costs: €756) and IL-6 (sensitivity: 0.70, specificity: 0.53, costs: €50) test were less cost-effective mainly due to higher costs of MRI (€756) and lower specificity of IL-6 (0.53). When using a QALY willingness-to-pay threshold of €30,000, genetic assay (sensitivity: 0.40, specificity: 0.85, costs: €750) is not cost-effective in any of the strategies. CONCLUSIONS: We have shown that our model is able to assess the cost-effectiveness of different diagnostic test strategies including the consequences for treatment decisions and disease course, and can easily be used for any new test for the early diagnosis of RA.