OBJECTIVE: This study tested the hypothesis that professionals' maintenance of long-term contact with persons who are at risk of suicide can exert a suicide-prevention influence. This influence was hypothesized to result from the development of a feeling of connectedness and to be most pertinent to high-risk individuals who refuse to remain in the health care system. METHODS: A total of 3,005 persons hospitalized because of a depressive or suicidal state, populations known to be at risk of subsequent suicide, were contacted 30 days after discharge about follow-up treatment. A total of 843 patients who had refused ongoing care were randomly divided into two groups; persons in one group were contacted by letter at least four times a year for five years. The other group-the control group-received no further contact. A follow-up procedure identified patients who died during the five-year contact period and during the subsequent ten years. Suicide rates in the contact and no-contact groups were compared. RESULTS: Patients in the contact group had a lower suicide rate in all five years of the study. Formal survival analyses revealed a significantly lower rate in the contact group (p=.04) for the first two years; differences in the rates gradually diminished, and by year 14 no differences between groups were observed. CONCLUSIONS: A systematic program of contact with persons who are at risk of suicide and who refuse to remain in the health care system appears to exert a significant preventive influence for at least two years. Diminution of the frequency of contact and discontinuation of contact appear to reduce and eventually eliminate this preventive influence.