OBJECTIVES: Delirium is common in hospitalised older patients with medical and surgical conditions, leads to poor outcomes and is an increasing burden on health care resources. A systematic review yielded no utility values for this population. The objective was to generate utility values to parameterise cost-utility decision models of delirium interventions in hospitalised elderly.  METHODS: A literature review identified only one study that: a) captured delirium diagnoses; b) employed a multi-attribute utility (or convertible) measure. The study included hip fracture and hip surgery patients in Sweden (n=115; mean age=83) and collected SF-36 data (admission and 6 months) and delirium diagnoses (n=32; 12.5% of these had delirium on discharge). The study author supplied the data and pre- and post-delirium utility values were calculated using i) SF-6D conversion ii) published EQ-5D mapping algorithm.  RESULTS: Pre-delirium utility differences between No Delirium and Delirium groups on admission were not in the expected direction but were non-significant. Mean(SD) SF-6D utility values were: Admission =0.598(.115) and 0.623(.081); 6 months =0.653(.144) and 0.618(.122), respectively for No Delirium and Delirium. Mapped EQ-5D scores were: Admission =0.544(.198) and 0.607(.136); 6 months =0.645(.224) and 0.586(.189), respectively for No Delirium and Delirium. Mean Admission-6 month changes were: (SF-6D) 0.055 and -0.005; (EQ-5D) 0.101 and -0.021 for No Delirium and Delirium, respectively. Thus according to both estimates, utility improved for those with no delirium but deteriorated for those who experienced delirium. CONCLUSIONS: Delirium has a lasting negative impact on patient utility. The utility values presented will be useful for future cost-utility analyses of delirium interventions targetting hip fracture. Assumptions are still required when populating health states and modelling lifetime impact as utility assessment during the delirium episode is not possible and longer term values are not available. Thus estimates at 1, 3 and post 6 months would provide a fuller picture of outcomes.