OBJECTIVES: Generic preference-based measures such as EQ-5D and SF-6D and condition specific measures such as the Diabetes Health Profile-18 (DHP-18) are used in the assessment of health related quality of life in diabetes. The aim of this study was to investigate the psychometric acceptability, validity and responsiveness of these measures in a type 2 diabetes population. METHODS: Acceptability was assessed by calculating missing data rates.  We investigated the convergent validity of the measures using correlations, known group validity across a variety of clinical indicators using ANOVA, and agreement between the measures using Intra Class Correlation coefficients (ICC) and Bland Altman plots.  Responsiveness to change over time was assessed using standardised response mean, and floor/ceiling effect tests. A UK longitudinal sample of 1613 people with Type 2 diabetes was used for the analysis. RESULTS: Overall rates of missing data were less than 5% indicating acceptability to respondents. Across both time points, the DHP-18 dimensions were moderately correlated with the EQ-5D and SF-6D indicating construct validity. All three measures significantly discriminated between those reporting and not reporting co-morbid health problems.  In terms of agreement, the ICC between the preference based measures was high, and Bland Altman plots indicated that agreement is higher at the less severe end of the scale. The EQ-5D displays evidence of a ceiling effect, and for patients self reporting change in health, SRM statistics indicate that the sensitivity to change across time is low for all measures. CONCLUSIONS: There is evidence for the psychometric validity of the generic preference based measures EQ-5D and SF-6D and the condition specific DHP-18 to measure outcomes in diabetes.  However evidence for the responsiveness of the measures is less clear.  The inclusion of both generic and condition specific measures in trials may increase the accuracy of outcomes assessment in type 2 diabetes.