A valid utility-based self-report measure is essential for evaluating health changes in employees. Health utility values permit cost utility analysis to be performed, an analytical technique rarely utilised in economic evaluations of workplace health promotion. There is no validated health utility measure in working populations. OBJECTIVES: To investigate the construct validity of SF-6D health utility instrument in a public service workforce. METHODS: The SF-12v2 Health Survey was administered to 3409 randomly selected public service employees in Australia in 2010. SF-12 scores were converted to SF-6D health utility values using Brazier’s algorithm. Associations with health, socio-demographic and work characteristics were explored using Spearman correlations. Linear regression was performed to identify correlates. Ceiling effects were analysed. For comparative purposes, normative SF-36v1 data from a general employed population were obtained from Release 11 of the Household, Income and Labour Dynamics in Australia (HILDA) survey (n=11234). RESULTS: Mean (SE) health utility was higher in males 0.794 (0.004, n=927) than females 0.773 (0.003, n=2387). Health utility was negatively correlated with psychological distress (K10) (r= -0.69 females, r=-0.69 males), comorbidity count (r= -0.35 females, r= -0.28 males), and effort-reward imbalance (ERI) (r= -0.36 females, r= -0.33 males). SF-6D was independent of age, BMI and annual salary. Lower SF-6D was significantly associated (p<0.05) with higher comorbidities, K10, ERI, education (in females), absenteeism, and lower physical activity (in females). The Australian normative mean (SE) SF-6D was 0.792 (0.002, n=4955) for males and 0.775 (0.003, n=4639) for females. Correlates and associations were similar in the normative sample, which showed an additional significant inverse association with age and positive association with salary in the general employed. CONCLUSIONS: SF-6D health utilities differentiated between and associated as expected with key health, social and economic factors. These results validate the suitability of SF-6D to measure perceived health states in an employed population.