Non-conventional monetary policy works in part because of what people believe. And these beliefs are not always fully rational in a textbook sense. This chapter describes two examples in which beliefs can affect outcomes. In the first example we explain how the effectiveness of forward guidance depends on whether people perceive it as an unconditional commitment that policy interest rates will stay low or as an adverse signal about economic prospects. Forward guidance can become counterproductive if more people perceive it as the latter than the former. In the second example we show that large-scale bond purchases by central banks can raise bond prices just because people believe they should. Both these examples show that behavioural finance has insights for monetary policy making.