OBJECTIVES: This study aims to investigate if the results of four published algorithms for calculating utility values from assessments of SF-36 are in agreement with the responses of traditional efficacy variables assessed in a randomized clinical study with different treatments of asthmatic patients. METHODS: Data from a randomized clinical study of moderate asthmatic patients comparing treatment with budesonide alone (n = 114) with budesonide plus formoterol (n = 109) during 12 weeks are used in this investigation. Utility values from the four algorithms are calculated for the different treatment groups at randomization and at end of treatment, and both absolute values as well as change during treatment are correlated with efficacy variables assessed in the study: PEF Morning, FEV1, and the summary score SF-36 PCS from the SF-36 questionnaire. RESULTS: Mean Utility values at baseline range between 0.61 to 0.82 for the 4 algorithms but with no difference between the two treatment groups. Change during treatment varies between 0.08 and 0.11. While both PEF Morning and FEV1 are statistically significant when comparing the change during treatment between the two treatment groups, neither any of the SF-36 domains nor SF-36 PCS turns out to be. Two out of the four utility measures, both based on TTO, reaches statistical significance. Correlation for change during treatment shows moderate correlation with PEF Morning (0.28 to 0.32) and FEV1 (0.17 to 0.25). CONCLUSION: The two utility measures based on the SF-36 items (or a subset thereof) and evaluated through TTO show better response than the other two, one evaluated through a Visual Analog Scale as rating scale, and the other based on domain values and not item values from SF-36.