OBJECTIVES: Payers are increasingly concerned with assessing the value of pharmaceutical products to guide reimbursement and pricing decisions, and particularly in countries employing an HTA process, value is often largely based on clinical effectiveness measured as cost-per-QALY. The objective of this study is to assess how the EQ-5D index reacts across a broad range of conditions to clinical changes which are considered significant at the individual disease and patient level. METHODS: A systematic review of published studies assessing the psychometric properties of the EQ-5D, in particular studies reporting statistics on the responsiveness of the EQ-5D index. The Standardised Response Mean (SRM) and Effect Size (ES), which are both measures of signal-to-noise and describe the responsiveness of EQ-5D to clinical change, are compared to benchmark scores from the psychometric literature. We also compare the Visual Analogue Scale (VAS) self-reported health component against the EQ-5D index. RESULTS: In the majority of studies, the responsiveness of EQ-5D to clinical change overall was poor to moderate: In only approx. half of studies was the VAS score within 10% of the EQ-5D index score. Ceiling effects of >15% at baseline where reported in half the studies examined. Only 25% of studies reported ES values within the range considered “moderate” to “good”, and 7% of studies reported SRM values in the “moderate” to “good” range. CONCLUSIONS: When considering widely accepted psychometric benchmarks and the relation between self-reported health and EQ-5D index, only a minority of studies (11%) report consistently good performance of the EQ-5D in reflecting clinically important changes. Consequently, a significant proportion of therapeutic value (both from a clinical and patient perspective) is not adequately captured by the existing EQ-5D framework. This has implications for the way new therapies have traditionally been assessed for value, particular as countries start moving towards value based pricing.