In this independent critique, based on four years of research, Stan Openshaw argues that reactor siting provides a simple means of offering additional, design-independent margins of safety. Reactor siting policies in the UK and USA are examined and it is suggested that UK siting practices need to be updated. The large number of potential alternative sites should be used to devise new planning strategies – strategies which will minimise both the residual health risks from accidents and the danger that a future change in public opinion might lead to calls for the closure of many existing sites on safety grounds.