OBJECTIVES To demonstrate how expert elicitation can be combined with Bayesian decision theory to inform the decision to conduct further research and identify efficient research design when there is limited evidence available. METHODS Using a model of enhanced external counterpulsation (EECP) for the treatment of angina, expert elicitation was used to provide informed priors on the key model parameters where no evidence was available. These priors were updated with sample information to quantify the expected value of sample information (EVSI). The EVSI was compared to the costs of sampling and the expected net benefit of sample information was obtained (ENBS). The ENBS is a measure of the societal payoff to research and provides a sufficient condition for deciding to conduct more research. Given the need to establish if EECP has a role for treating angina and the paucity of effectiveness evidence available, the ENBS was used to determine the type of study required, the optimal sample size, the appropriate length of follow-up and the endpoints that should be included. RESULTS There is significant value to future research in EECP. This research is most valuable when directed towards establishing the short- and longer-term effectiveness of EECP. Although the ENBS depends on the cost-effectiveness threshold, the ENBS is positive for most sample sizes indicating that further experimental research will be efficient. A clinical trial design is proposed with equal allocation of patients between arms. Given a threshold of ≤20,000 per QALY gained, the highest expected returns to additional research will come from a trial with a 4-year follow-up and an optimal sample size of 900. CONCLUSIONS Expert elicitation and value of sample information can be combined into a single coherent framework to establish the value of further research, the optimal design of such research, and the most appropriate basis for informing cost-effectiveness.