OBJECTIVES: Health state utility values (HSUVs) are required for cost-effectiveness analysis of new medicines. Health-related quality-of-life data collected from the widely accepted EQ-5D are often not available and data from other instruments need to be mapped using algorithms. The objective of this study was to map HSUVs using data from a clinical trial of patients with relapsed/refractory multiple myeloma (RRMM).  METHODS: Patient-level EORTC QLQ-C30 and EORTC QLQ-MY20 data were collected in a clinical trial of RRMM patients (n=640) and mapped to EQ-5D scores using published algorithms. Descriptive summaries of mapped EQ-5D scores were estimated overall, and treatment regimen by: health state (pre-progression, progressive disease), response (with/without complete response), and adverse events (with/without grade ≥3 AEs). Generalized estimating equation (GEE) models provided direct estimation of HSUVs, controlling for confounders. Mapped HSUVs were compared with published HSUVs in RRMM to assess reliability.  RESULTS: The algorithm that included the EORTC QLQ-C30 and EORTC QLQ-MY20 produced reliable HSUV estimates providing greater differentiation between health states. The overall mean estimate for progression-free disease (PFD) was 0.733. PFD with response was 0.744 and PFD with no response was 0.704. The difference between scores with/without grade 3 AEs suggested a utility loss of 0.029; the AE decrement was greater (0.034) when using a proxy measure of patients who were off treatment but not in progressive disease. Similar patterns across the health states were seen by treatment regimen throughout the duration of the trial. GEE results were consistent with descriptive summaries.  CONCLUSIONS: Mapped EQ-5D scores showed a consistent trend across health states, with higher HSUVs for pre-progression than progressive disease, for response versus no response, and without AEs versus with AEs. HSUVs were dependent on disease state and treatment regimen. The algorithm including both the EORTC QLQ-C30 and EORTC QLQ-MY20 provided reliable HSUV estimates in this RRMM population.