Autopsies continue to be the most reliable source of mortality statistics; however, more and more death certificates are based on the post-mortem external examination (PME) alone. Forensic PMEs differ from clinical PMEs, because the forensic pathologist usually has no preceding knowledge of the health of the decedent and must rely on information from authorities in the form of the police report. It is useful at the forensic PME to know whether the decedent suffered from a mental illness; however, it is unknown how valid such a diagnosis is, when based upon information in the police report alone. This study compared tentative diagnoses of schizophrenia from 500 forensic PMEs with a reference database based on the Danish National Patient Registry. We found that 19.3% of schizophrenia cases were missed, and 9.1 % of identified cases were false positives. Overall, 11.4% of all assessments were incorrect. Subgroup analysis showed that marital status as 'single' and the finding of illegal substances at the scene were predictors for both correctly identified and overlooked schizophrenia cases. The most reliable source of information was the decedent's general practitioner, whereas friends and neighbors were the most unreliable. Future studies should be aware of the risk of assigning a wrong diagnosis and use as many sources of information as possible. Taking the decedent's social history and observations about the scene into account may add to the diagnostic accuracy.