Discounting is a technique commonly used in cost-effectiveness analysis to 'make fair' comparisons of programmes whose costs and outcomes occur at different times. It is not a correction for inflation. While there is general agreement among health economists regarding the need to discount, there is less consensus on the procedure for discounting costs or benefits. We describe the method of constant-rate discounting, which uses the same rate to discount costs and benefits, and examine its impact on the cost effectiveness of selected health interventions. This methodology has significant limitations, however. Constant-rate discounting may not accurately represent the values of a society. Furthermore, discounting does not reflect the longitudinal time preferences of individuals (or groups). The philosophical rationale for constant-rate discounting is the concept of longitudinal equity, i.e. that society should make allocation decisions in such a way that present and future cohorts are treated equally, regardless of when they come into existence. In general, discounting affects the cost effectiveness of preventive interventions more than acute interventions, and it affects programmes with immediate cost more that those with ongoing cost. The reader of such analyses should be aware of these effects, and should use caution in comparing the cost effectiveness of interventions with vastly different timing of cost and benefit.