The psychosocial implications of cancer and its treatments are well recognized. In NSCLC, many patients are diagnosed at an advanced stage with substantial impairment of QoL. While psychometric-based measures allow the assessment of patients in a descriptive profile of one or more QoL domains, utility measures allow the expression of the overall experience of QoL in a single number that represents patient preferences for different levels of QoL. However, the scientific validity of QoL and utility assessments is determined by the reliability and validity of the measures used. OBJECTIVES: To identify utility and psychometric-based instruments used in NSCLC and compare evidence in support of their validity, reliability and responsiveness. METHODS: We conducted a search of MEDLINE, EMBASE and within our Information Resources Centre from 1985 to date using various terms including NSCLC, QoL and utilities. RESULTS: Instruments were selected for review if they had been developed or used in a study population of which at least 20% were NSCLC patients. The search and selection yielded 26 original studies in which 4 utility measures (TTO, TrTO, HUI and EQ-5D) and 12 psychometric-based health profiles such as the SF-36, FACT-G and the EORTC-QLQ were applied. Our review of published data revealed striking differences between measures in terms of the coverage of QoL conceptual domains and psychometric properties. Compared to health profiles, relatively scant evidence was found in support of the psychometric validity of utility measures. CONCLUSION: Findings from this review suggest a need for further validation of existing utility measures in order to enhance the collection of scientifically valid preference-weighted QoL data during studies of patients with NSCLC.