Background Psychological autopsy studies consistently report that the rate of detected mental disorders among suicide decedents is below 100%. This implies three possibilities: (a) a subset of suicide decedents did not have a mental disorder at the time of death; (b) all suicide decedents suffered from a mental disorder, but some were undetected due to methodological limitations; and/or (c) suicide decedents with an undetected mental disorder displayed significant and perhaps subclinical features of a mental disorder. Objective In this article, we examined these possibilities by evaluating the differences in symptoms and stressors between suicide decedents who were undiagnosed and those diagnosed with a mental disorder at the time of death. Method We reviewed 130 case studies of community‐based suicide decedents originally described in Robins’ (1981) psychological autopsy study. Results Without exception, suicide decedents in Robins’ sample suffered either from a clearly diagnosable mental disorder or displayed features indicative of a significant, even if subclinical, presentation of a mental disorder. Undiagnosed and diagnosed suicide decedents did not significantly differ with regards to demographics, violence of suicide method, suicide attempt history, the number and intensity of stressful life events preceding death, and whether their death was a murder‐suicide. Conclusion Although clearly not all who suffer from mental disorders will die by suicide, these findings imply that all who die by suicide appear to exhibit, at minimum, subclinical psychiatric symptoms with the great majority showing prominent clinical symptoms. We conclude with clinical implications and recommendations for future study.