Policies aimed at preventing (further) radicalisation or aimed at de-radicalisation are required to be 'evidence-based'. This suggests that evaluators should apply rigorous empirical methodology and measurement techniques. However, it is often unclear what this evidence should consist of and how it should be gathered. In the present paper we present results of a literature review focusing on evaluations of programmes aimed at preventing radicalisation or de-radicalisation between 1990 until July 2014. We identified 55 manuscripts including 135 participant samples. Primary qualitative or quantitative empirical data about effectiveness of an intervention was presented in only 16 participant samples (12%). The outcomes are discussed with respect to methods and interventions used in the research field of criminology, a valuable source of methodological experience in conducting evaluation research in challenging circumstances. We recommend the use of empirical studies using quantitative data when possible (i.e., in preventive interventions) and a multi-method approach for evaluating programmes in (even) more challenging contexts (i.e., de-radicalisation programmes).