The recent economic evaluation of an Improving Access to Psychological Therapies (IAPT) service conducted by Mukuria and colleagues ^{Reference Mukuria, Brazier, Barkham, Connell, Hardy and Hutten1} is a welcome addition to the evidence base pertaining to this programme. This was a non-randomised comparison but it appears that the authors have used appropriate methods to control for differences between areas. A casual reading of the abstract conclusion would lead one to assume that IAPT is likely to be cost-effective. Indeed, the cost per quality-adjusted life year (QALY) is below the upper threshold used by the National Institute for Health and Clinical Excellence (NICE), and below the lower threshold in a sensitivity analysis where the EQ-5D was used. However, the cost per QALY is somewhat misleading. The most useful results from this study are the cost-effectiveness acceptability curves shown in Fig. 2. Here it is revealed that at the NICE upper threshold of £30 000 per QALY, there is about a 38% likelihood that IAPT is cost-effective, increasing to just over 50% if the EQ-5D is used to generate QALYs. If the lower threshold is used, then there is even less chance that IAPT is cost-effective. The overall conclusion of this paper should be based on Fig. 2 and it should be that on the basis of this study IAPT was probably not cost-effective.