OBJECTIVES: In order to construct a cost-utility model, evidence is required of the Health Related Quality of Life (HRQL) experienced by patients suffering from the disease. In advanced melanoma, data is available from a vignette-based standard gamble (SG) study in the general population. However, patient reported HRQL data was also captured in the ipilimumab pivotal trial, MDX010-20, using the EORTC QLQ-C30 and the SF-36 generic health surveys. METHODS: Patient level EORTC data from the MDX010-20 trial was mapped using the EORTC-8D algorithm to produce EQ-5D utilities, which were then stratified according to disease progression (progression-free  or post-progression), and treatment arm. This process was repeated with the patient level SF-36 data based upon a nonparametric Bayesian method to generate SF-6D utility values. The results were then compared with results generated from the vignette-based study. RESULTS: In the progression free health state, the SG and EORTC data show a high degree of correlation in utility (0.77 vs 0.80), with the SF-36 value being significantly lower (0.64). In the post-progression state, comparing to the utility values in the progression free state, the SG data shows a significant fall of 0.18 (23.4%) in expected utility, however, this is not mirrored in patient data, where there is a fall of 0.04 (4.7%) in the EORTC data, and 0.02 (3.3%) in the SF-36 data, showing patients do not appear to have a significantly worsened HRQL with disease progression. CONCLUSIONS: Despite the limitations of the study in both patient numbers, and being limited to a single disease, investigators should be aware different measures administered to the same patients may yield differing results. Equally further research should be carried out on HRQL associated with disease progression from the viewpoint of both patients and the general public, as it is possible there are differences in the valuation of states close to death.