Evaluation of quantitative easing (QE) is difficult as it is only used in response to severe and unusual economic difficulties. Despite this, we argue that two main conclusions can be drawn from a sceptical reading of the evidence. First, large-scale asset purchases reduce government bond rates, especially at the longer end of the yield curve. However, this effect may be temporary and is small if bond rates are already low, while initial waves of QE are more effective than subsequent programmes. Second, QE appears to have been effective in late 2008 and 2009, preventing even larger declines in output and inflation than were experienced. We argue that the literature is limited, relying on similar methodologies and largely originating in central banks. Exploration of alternative approaches to QE would be useful in widening an evidence base that is currently too narrow.