OBJECTIVES: The arrival of personalized medicine has led to an increasing recognition of the need to reconcile the evidence requirements for tests and treatment technologies. In this paper we build on Phelps and Mushlin’s (1988) model for optimizing diagnostic technologies and develop an analytical framework to assess the cost effectiveness of personalized medicine co-dependent technologies. METHODS: The co-dependent technology that we analyse consists of two tests and a treatment decision guided by the tests. The first test is a dichotomous test which identifies the existence of a health condition, the second test characterises the magnitude of response to therapy. We consider three treatment choices; new treatment for patients who benefit from therapy; alternative treatment for patients who do not benefit; no treatment. Within our framework, an individual’s response to therapy is exogenously determined and a payer decides on a cut-off for the second test such that patients whose test score lie above this cut-off are deemed to benefit from therapy and hence are administered the new treatment. Individuals with test scores below the cut-off receive standard care. We investigate the optimal decisions by the payer under different settings and determine how the distributions for probability of response and benefits influence the net health benefits acquired by the payer. RESULTS: Our preliminary results suggest that for a given probability of response, the payer sets the optimal test score at the point where the marginal benefit of the new treatment is equal to the marginal benefit of standard care. At this optimal point the payer is indifferent between funding the new treatment or standard care. CONCLUSIONS: Obtaining value from the use of co-dependent technologies requires taking into consideration the joint dependence of all test and treatment technologies.