Background: Risk assessment and management of suicidal patients is emphasized as a key component of care in specialist mental health services, but these issues are relatively unexplored in primary care services. Aims: To examine risk assessment and management in primary and secondary care in a clinical sample of individuals who were in contact with mental health services and died by suicide. Method: Data collection from clinical proformas, case records, and semistructured face-to-face interviews with general practitioners. Results: Primary and secondary care data were available for 198 of the 336 cases (59%). The overall agreement in the rating of risk between services was poor (overall κ = .127, p = .10). Depression, care setting (after discharge), suicidal ideation at last contact, and a history of self-harm were associated with a rating of higher risk. Suicide prevention policies were available in 25% of primary care practices, and 33% of staff received training in suicide risk assessments. Conclusion: Risk is difficult to predict, but the variation in risk assessment between professional groups may reflect poor communication. Further research is required to understand this. There appears to be a relative lack of suicide risk assessment training in primary care.