OBJECTIVES: To derive utilities for adults with ADHD from a randomised trial of atomoxetine 40mg BID versus 80mg QD (Adler et al., 2004). These patients are largely untreated and there are few therapy alternatives. METHODS: Medical Outcome Study Short Form-36 (SF-36) scores are converted into a Quality Adjusted Life Year (QALY) value for key health states; responder and non-responders without/with adverse event (grouping defined as per clinical trial). The method for conversion is in accordance with that previously published (Brazier et al. 1998, 2004). A total of 218 clinical trial participants were followed for 13 weeks. Pre-treatment utility data of all patients is analysed to assess baseline utility values and is compared with end of trial period. Missing data were addressed according to five criteria. RESULTS: Mean and median results for utility values for responder and non-responder groups produced consistent and sensible results (mean results presented unless stated). Baseline utility was 0.634 and was consistent with non-responder score at end of trial (0.630), as per expectation. Responders had utility scores of 0.682 and 0.671, without/with adverse events, respectively. All responders (i.e. regardless of adverse event status) had a score of 0.678. The average gain in utility at end of trial comparing non-responders with all responders was 0.048. Median results increased the utility gain to 0.08. CONCLUSIONS: This analysis provided results that were rational and consistent with efficacy and safety findings with atomoxetine in the treatment of adults with ADHD. Improved utility can be expected with treatment (0.048 to 0.08) and these data can be used to populate an appropriate cost-utility analysis. Conversion of SF-36 data to SF-6D values was informative and consistent.