The research examined deaths for all municipal employees and police officers who worked at last 5 years in their specific occupational group between 1950 and 1990. The reports of the 139 deaths that were officially classified as suicide, accidental and undetermined were submitted to a panel of medical examiners for validation. Results revealed that six cases that were originally listed as accidents or undetermined were reclassified as suicide. Findings indicated that official police suicide rates were less sensitive than the rates of other workers in detecting actual suicides and had less predictive value in determining actual nonsuicides. Findings have implications for both police training and assistance to survivors of police suicides. Tables and 41 references (Author abstract modified)