OBJECTIVE: To investigate if a health-related quality of life (HRQOL) measure can predict pharmacy costs in insured adult asthmatics independently and in conjunction with a baseline pharmacy claims risk index and prior pharmacy costs. METHODS: Historical cohort analysis of pharmacy claims and HRQOL. Asthmatics were invited to complete the SF-36 quality of life instrument between May 12, 2001 and November 30, 2002. Pharmacy claims for the year prior to this interview were used to derive baseline measures of the AdvancePCS pharmacy claims health risk index and prior pharmacy costs. The study's outcome measure, pharmacy costs in the year following the baseline interview, was calculated for those who remained eligible for pharmacy benefits during the entire study period. Correlation statistics were calculated. Linear regression modeling evaluated the association between the outcome measure and each SF-36 domain, the risk index, prior costs, member sex, and age at baseline. RESULTS: A population of 247 adult asthmatics from a million-member, Mid-Atlantic health plan was included. Follow-up year pharmacy costs were moderately correlated with the pharmacy risk index (Spearman + 0.576, p <0.001) and weakly negatively correlated with the physical summary score (Spearman -0.340, p <0.001). However, prior pharmacy costs had the strongest correlation (Spearman + 0.87, p <0.001). The physical summary score was a significant predictor of next year's pharmacy costs (Coefficient -110.92; p <0.001) after adjustment for age in linear models (R2 = 0.134). However including the pharmacy risk index (Coefficient 72.770; p = 0.06) and prior pharmacy costs (Coefficient 0.948; p <0.001) vastly improved the explanation of variation in follow-up pharmacy costs (R2 = 0.774) and eliminated the independent significant effect of the physical summary score (Coefficient -8.236, p = 0.518). Similar removal of significance occurred with other SF-36 domains after adjusting for claims measures. CONCLUSIONS: While HRQOL independently predicts follow-up costs, it does not improve predictive models using prior pharmacy utilization in this study population of adult asthmatics.