OBJECTIVES: Utility calculation generally relies on the use of generic measures that assess health status. Evidence suggests that such measures have limited responsiveness - largely because the health states described are of limited relevance to any specific disease. Quality of life (QoL) assessment has shown that it is not health status itself that is important to patients but the impact that this might have on their lives. Consequently, it would appear more appropriate for utility to be determined by preference for different QoL states, particularly in Quality Adjusted Life Year (QALY) type analyses. The present study was designed to determine QoL-based utilities specific to RGH. METHODS: Discrete choice conjoint analysis (CA) and time trade off (TTO) exercises were conducted using QoL states generated from six items from the RGH Quality of Life Questionnaire (RGHQoL). RGH patients completed tasks via interview. RESULTS: One hundred ninety-two interviews were conducted (79 male, 113 female; age range 19-69 years, mean age 38.4). For CA, all attributes were statistically significantly influential in determining preferences. The application of the random effects probit model produced coefficient values that can be used as preference weights. For the TTO exercise, coefficient values (preference weights) were derived by application of the random effects tobit regression model. These coefficient values can be used to derive relative and absolute utility values respectively. As the TTO technique possesses cardinal properties, QALY scores can also be calculated. CONCLUSIONS: It is feasible to generate both relative and absolute utility values from responses to the RGHQoL questionnaire, allowing utility to be based on true QoL. The ability to derive disease-specific QoL-based utilities in this way means that the same instrument can be used to generate both QoL and utility data from the same clinical trial.