This paper is based upon an extensive review of 71 willingness-to-pay (WTP) surveys of health and health care published in English during the period 1985--1998. The aim of the paper is to outline the arguments advanced for the superiority of WTP over quality-adjusted-life-years (QALYs) as a measure of benefit of health care programmes, and to review how empirical WTP studies adhere to their implications. An important argument is that WTP enables a more comprehensive valuation of benefits than QALYs. Our main focus is therefore to provide a careful review of the scenario descriptions used in the surveys, according to which types of benefits are being valued, and how comprehensively the descriptions are presented. Furthermore, the 'cost-benefit argument', that WTP can assist in improving social efficiency, is discussed before we inquire into the extent to which the studies actually compare WTP with social costs.