In population studies, many risk factors are associated with suicide completion. Yet we cannot accurately predict whether any individual patient will die by suicide. Completers are a distinct population from attempters and do not necessarily present for treatment by mental health professionals. Research on suicide prevention has yielded some promising findings but has not shown that interventions produce definitive results. The strongest evidence for successful prevention derives from reducing access to means. A population-based strategy may be more effective than a high-risk strategy focusing on patients with suicidal ideas or attempts. Much more research is needed before developing effective suicide prevention programs.