The increasing frequency of suicide in Norway from 1970 to 1990 has called attention. This paper raises the question of whether any controlled studies have been conducted on prevention of suicide. A search was made in Medline; the criteria for including articles were: controlled studies concerning the effects of suicide-preventive intervention by measuring the number of suicides, parasuicides or suicidal thoughts/ideation. The result of our search was 13 randomised controlled trials and two case-control studies. Most of the studies were unable to confirm any effect of suicide-preventive intervention. In most studies the experiment and control groups were far too small to arrive at significant conclusions. Prior to starting prevention programmes, it is necessary for the health service to know the potential effects. This systematic review revealed that few good controlled studies have been conducted on suicide prevention. For the time being we have to accept the suggestive evidence for certain types of intervention, and use these as a basis for action.