Between 1999 and 2019, according to the Centers for Disease Control and Prevention (CDC), nearly 450,000 people died from overdoses involving prescribed opioids. This review examines how drug overdose deaths are compiled by the CDC using the coding system of the International Classification of Diseases (ICD). When it comes to drug-involved deaths, the ICD may not tell the whole story or even the right story. To learn why, the authors examined the CDC's source data and the standard death certificate. In fatal drug overdose cases, death certificates are issued often before the results of post-mortem toxicology are known by the certifier. The CDC believes that this delay in the death investigation process may account for errors when, for example, certifiers list ambiguous terms such as "suspected acute drug intoxication" or "possible drug overdose" as a cause of death. When incomplete data are coded according to the ICD, the error is passed along while potentially useful information is lost. The result may reflect accurately the annual total of drug-involved overdose deaths while obscuring the lethality of individual substances, consumed alone or in combination, which contributed to, or caused, drug-involved deaths. The true cause of most fatal drug overdoses-polysubstance abuse-often is lost in this process. A key objective of this paper is to describe the process used by the CDC to report drug-involved mortality and how the current iteration of the ICD may be ill-suited for this important task.