OBJECTIVES: Discrete choice experiments (DCEs) are choice-based methods used to estimate utilities. While DCEs are being used increasingly in health economics, their application to estimating utilities for health states and outcomes is more novel. Visual analog scaling (VAS) is a well-established rating method used widely in health care. Previous studies have shown that DCE and VAS results are highly correlated. This study compares state and attribute utilities for Barthel Index (BI) physical disability states estimated by DCEs and the VAS method. METHODS: A convenience sample of 152 subjects (= 45 years) able to answer the 37-question survey was enrolled. Demographic characteristics, and physical disability and caregiving histories were collected. The DCE consisted of 13 choice sets. The VAS exercise used a rating thermometer (0 = worst imaginable health state, 20 = best imaginable health state) to rate 14 scenarios. The survey was administered in individual and small group settings by trained researchers. Utilities for disability states and attributes were analyzed using ordinary least squares and probit regressions. Spearman correlations were calculated to compare utilities and BI scores. RESULTS: Thirty-one percent of the subjects were male, mean age was 61 years, 88% were Caucasian, 60% had more than a high school education, and 63% were married. The strongest determinants of VAS utilities were Feeding (p = .002), Continence (p = .003), Mobility (p < .001), and Bathroom Use (p = .041). Feeding, Bathroom Use and Personal Care were the strongest determinants of DCE utilities (p < .001). The relative contribution of Feeding was strongest in the DCE model. Correlations between VAS and DCE utilities, and BI scores were 96% and 91%, respectively. The correlation between VAS and DCE utilities was 94%. CONCLUSIONS: Although highly correlated, attribute contributions to DCE and VAS utilities differed. Subgroup analyses and exploration of design issues should help clarify the basis for the model differences.