OBJECTIVE: Discrete choice modelling (DCM) applied to data generated by stated preference (SP) experiments is being used increasingly by health economists as a method for valuing process and outcome characteristics of health interventions in utility or monetary terms. The purpose of this paper is to illustrate how sponsors of randomised controlled trials (RCTs) could use DCM to assist with the planning of their studies. The approach is illustrated using a case study of the design of trials to evaluate the use of adjuvant bisphosphonates in the management of patients with primary operable breast cancer. METHODS: A stated preference experiment was conducted amongst UK specialists involved in the management of patients with primary operable breast cancer. Respondents were asked to choose 1 bisphosphonate regimen from each of 16 binary choice scenarios. Each regimen was characterised in terms of the following trial design attributes: i) Primary endpoint ii) Effect size demonstrated iii) Uncertainty surrounding demonstrated effect iv) Duration of observation v) Study population vi) Cost of the treatment alternatives. The survey was performed using a telephone-mail-telephone approach. Probit analysis was used to estimate a binary choice model of drug prescribing behaviour. RESULTS: 54 specialists fully completed the survey questionnaire providing a sample of 864 discrete choice responses. In qualitative terms, the signs on the coefficients were in line with prior expectations and all coefficients were statistically significant at conventional levels. The marginal and average effects were used to determine the relative importance of the attributes and to rank alternative designs in terms of the ex ante probabilities of product adoption. CONCLUSIONS: DCM could be used by sponsors of RCTs to incorporate decision-maker preferences into their designs. It could also be used to estimate product uptake contingent upon different designs. Results from this study suggest that the approach is both feasible and valid.