AbstractBereaved people are at greater risk of losing their lives than the non‐bereaved. This paper explores the evidence on this ‘bereavement—mortality relationship’, to establish who among the bereaved are most vulnerable to such dire consequences of loss. Information is drawn from several sources, including epidemiological research on sociodemographic variables, analyses of causes of death and recent empirical studies that have identified differences between bereaved decedents and survivors. The survey shows that risk factors interact in complex ways to determine level of vulnerability. A simple explanation in terms of ‘broken heart’ cannot account for the pattern of results. Some bereaved subgroups are more vulnerable to death than others, excess risk occurs across a variety of causes of death (suggesting different underlying processes), and personal characteristics in combination with circumstantial factors lead to high vulnerability. It is concluded that two types of process underlie the bereavement–mortality relationship. These are related: (a) to the direct effects of grief (broken heart); and (b) to secondary consequences associated with the stress o