OBJECTIVES: Health state utility values (HSUVs) for non-small cell lung cancer (NSCLC) vary by instrument and valuation method, as well as by patient and disease characteristics such as histology and mutation status. This research aimed to identify the range of HSUVs implemented in economic models for advanced NSCLC in the United Kingdom. METHODS: A targeted review of NICE health technology appraisals (HTA) was undertaken to identify HSUVs in patients with advanced NSCLC. Values generated through mapping were excluded. RESULTS: Fifteen submissions were identified reporting HSUVs for progression-free (PF) or progressed disease (PD). These submissions applied HSUVs derived from 11 trials and 3 valuation studies, and they tended to focus on chemotherapies rather than oral anti-cancer therapies. The median (range) values identified were 0.706 (0.620, 0.815) for PF and 0.592 (0.470, 0.688) for PD. Most appraisals reported HSUVs calculated as simple averages across EQ-5D index scores pre- and post-progression. Only a few used time-varying values within PD or prior to death, or values adjusting for baseline characteristics. Trial based HSUVs were 0.740 (0.687, 0.815) for PF and 0.640 (0.550, 0.698) for PD. Non-trial based studies generally reported lower values: 0.650 (0.620, 0.659) for PF and 0.473 (0.470, 0.530) for PD. The lowest values were found in a study in which HSUVs were derived from hypothetical health states administered to the general population using a standard gamble. CONCLUSIONS: There are large variations in HSUVs derived from studies for advanced NSCLC. Trial based HSUVs elicited from the patients being investigated are the most relevant and aligned with NICE’s reference case. Non-trial based HSUVs represent an alternative source but given the wide range of reported values it is important that they are consistent with the patient population and treatment profile, and that their use is appropriately justified, especially where trial based values exist.