OBJECTIVES: NSCLC (accounting for 80% of all cases of lung cancer) causes high morbidity and mortality. Various treatment lines are available for NSCLC, and there are ongoing discussions on the most appropriate measure of treatment efficacy for reimbursement decisions. We examined the use of these endpoints in NSCLC from a payers’ perspective.  METHODS: Targeted searches were conducted in MEDLINE® and the Cochrane Database of Reviews, using clinical and health economic-related key words and limited from 2000 to 2010. RESULTS: OS can be measured easily and accurately in terms of both event and time and is the endpoint preferred by regulatory bodies including the FDA and EMA. Over the last 10 years the benchmark OS for first-line treatment of NSCLC has risen and it is increasingly difficult to demonstrate significant OS benefit, as the efficacy (clinical trial results) of treatments has improved. Demonstrating improvements in OS over best supportive care is less challenging than against an active comparator. Adequate powering of studies to demonstrate OS benefit is vital but requires large study populations. Demonstrating a significant PFS benefit is also challenging, requiring frequent assessments, precise event definition and exact determination of the time of event. Nevertheless, PFS provides a well-accepted alternative endpoint to OS as it is a direct measure of treatment effect on tumour burden and measures only the effect of the study drug. PFS has also been accepted by regulatory bodies as a measure of the efficacy of cancer treatments. CONCLUSIONS: OS is generally regarded as the preferable endpoint (from a payer’s perspective) for demonstrating clinical efficacy in NSCLC. There are challenges, however, with demonstrating OS benefit of new therapies for NSCLC. PFS data may be more appropriate for use in certain situations, especially those in which subsequent lines of therapy exist.