Several trials have employed anxiety measures to assess decision aid effectiveness. This study employed a systematic review method to integrate their findings. The affective impact of decision aids and the appropriateness of anxiety as a measure of decision aid effectiveness are explored. From 11,361 citations generated from searching electronic databases and journals, 26 randomised controlled trials evaluated patient decision aids; 10 included anxiety measures (HADS; STAI). The data were too heterogeneous to integrate statistically. No studies showed an increase in anxiety from exposure to decision aids versus usual care. Some patterns emerged between level of anxiety and characteristics of the decision. As raised levels of anxiety are associated with both more effective decision strategies and stressful health interventions, anxiety is an inappropriate measure to employ when evaluating decision aids. Future research needs to investigate the relationship between affect, cognition and decision aids in order to facilitate effective patient decision making.