Since the mid‐1980s, ‘indicators’ have been increasingly suggested as a possible means to evaluate the effectiveness of various policy mechanisms. In this paper, we shed some critical light on the use of indicators, by investigating in detail the advantages and disadvantages of using the indicator approach in the assessment of agri‐environmental policy effectiveness in the European Union. We review the theoretical and conceptual bases of the indicator approach, highlighting the complexity of identifying appropriate indicators and the difficulties in identifying means to ‘measure’ them. We analyse four levels of indicators (uptake, output, actor‐related and procedural indicators), and argue that, while indicators continue to have a central role to play in the evaluation of European agri‐environmental policy, that role is essentially limited and the notion of what constitutes an indicator needs to be thoroughly re‐examined. Copyright © 2001 John Wiley & Sons, Ltd and ERP Environment.