A companion diagnostic test may be used to identify patients with certain biological traits and guide optimized therapies to these patients. There is a great deal of interest in developing companion diagnostics for therapies to improve efficacy and minimize harmful side effects. There are many statistical issues associated with companion diagnositics including the exploratory analyses methods in the earlier phases of trials to establish the discriminative cutpoint and clinical trial design to validate the companion diagnostic test as a patient selection device for the therapy. The solutions to these problems depend on whether the test is simultaneously under development with the therapy or an established test, whether the test is a lab developed test or a market ready test. The roundtable provides a platform to discuss the various issues associated with companion diagnostics.