OBJECTIVES: The timely diagnosis of deep venous thrombosis (DVT) is critical because this disorder can be life threatening. However, referring all patients suspected of DVT for ultrasound (US) testing is inefficient since 80 to 90% of those referred have no DVT. Therefore, we investigated the cost-effectiveness of a diagnostic strategy based on a point of care d-dimer test combined with a clinical decision rule that was documented to be safe in primary care (AMUSE study). METHODS: A model based cost-effectiveness analysis was conducted in conjunction with a recent multi centre prospective diagnostic study (AMUSE, N=1002). A Markov model with a five year time horizon was used to compare the AMUSE strategy to two hospital based strategies: ultrasound for all and a hospital decision rule. Probabilities were derived from AMUSE and the literature. Societal costs and health state utilities were used. One way and probabilistic sensitivity analyses were conducted. Cost-effectiveness acceptability curves were constructed. RESULTS: The AMUSE strategy has both slightly lower costs and less quality adjusted life years (QALYs) than both hospital based strategies. The ultrasound for all strategy has the highest costs and QALYs, but is not cost-effective as compared the hospital decision rule strategy. The AMUSE strategy compared to the hospital decision rule strategy resulted in a mean saving of €138, and a mean QALY loss of 0.002. The incremental cost-effectiveness ratio is €56,436 per QALY lost. The cost-effectiveness acceptability curves show that the AMUSE strategy has the highest probability of being cost-effective, even exceeding ceiling ratios of €80,000 per QALY.  CONCLUSIONS:  The AMUSE strategy to exclude DVT in primary care is not only safe, but also has the highest probability of being cost-effective as compared to hospital based strategies to diagnose DVT.