To date, there has been little examination of the problems associated with conducting economic evaluation for interventions designed to contain antimicrobial resistance. There are two quite different types of intervention aimed at containing antimicrobial resistance: interventions which are designed to avoid the emergence of resistant organisms; and interventions that are designed to avoid the transmission of resistance organisms. Four aspects of economic evaluation where the ease of assessment might be expected to differ across evaluations for these different types of intervention are examined: problems associated with the identification of diffuse impacts, problems associated with comparing current and future impacts, problems associated with uncertainty, and problems associated with difficulties in measurement and valuation. The paper suggests that it may be much easier to conduct rigorous economic evaluations for interventions designed to avoid transmission of resistance, than for those intended to avoid emergence. Unfortunately, the transmission policies, which are likely to be the easiest to evaluate, are not likely to produce an optimal long-term outcome given the apparent irreversibility of much resistance and the potentially severe harms which could be imposed as a result. Given the desirability of avoiding a scenario where, in the evidence-based medicine culture, the most rigorously evaluated policies are followed even though they may be less important, there is the need to consider carefully what, and how, economic evaluation should be conducted in the area of antimicrobial resistance. It is suggested that research should focus on the use of modelling as a means of evaluating optimal policy responses and on trying to resolve some of the difficulties associated with measurement and valuation.