Background and Objectives: Historically, the duty of the medical examiner in assigning cause and manner of death in drug-related death cases has been fraught with controversial challenges. The lack of standardization in certifying drug-related deaths may involve differences among practicing forensic pathologists in their approach to such cases. The central objectives of the present study include characterization of current drug death patterns and the variability among medical examiners with respect to autopsy performance and death certification practices in one county medical examiner's office., Materials and Methods: Death certificates, scene information/investigative reports, autopsy reports, and toxicological laboratory results for each of the 100 cases of drug-related death occurring in 2002 in Fulton County, Georgia were reviewed. Comparison of overall autopsy rates and autopsy rates in drug-related death cases for each medical examiner individually and for the group collectively was performed. In examining cocaine-related deaths (most common), statistical analysis was performed for comparison of drug concentrations (cocaine and benzoylecgonine) between deaths certified as cocaine toxicity (poisoning) versus cocaine-complicating disease or causing an adverse event such as cerebral hemorrhage., Results: Causes of accidental drug deaths included cocaine 40%, mixed drug intoxication 37%, opioids 10%, ethanol 7%, and prescription medication (nonopioid) 5%. Overall total autopsy rates in 2002 for each of the 6 independent medical examiners ranged from 51% to 69% (mean 64%), whereas autopsy rates in drug-related death ranged from 55% to 91% (mean 81%). In review of the subset of 40 cocaine-related deaths, 25% were certified as cocaine toxicity (poisoning), with the remaining 75% certified as cocaine-complicating disease or causing and adverse event. Autopsy rates in cocaine-related deaths were as follows: cocaine toxicity 80%, cocaine-complicating disease 77.3%, and cocaine causing adverse event 62.5%. Thirty-eight percent of cocaine-related deaths were considered to be of "low suspicion" for drug involvement at the time the death was reported to the medical examiner with the remaining 62% being of "high suspicion". Autopsy rates were somewhat lower in the low suspicion group (67%) versus the high suspicion group (72%). Comparison of drug levels between cocaine-related death certification groups was performed. No statistically significant difference was shown in drug levels (cocaine, P > 0.3; benzoylecgonine, P > 0.2) between deaths certified as cocaine toxicity versus those certified as cocaine-complicating disease or causing adverse event., Conclusions: In Fulton County, accidental drug deaths in 2002 most often involved cocaine either alone or in combination with opiates and/or alcohol. Cocaine, opiates, or both were involved in greater than three-fourths (77%) of all drug-related deaths. The majority of all decedents were black (57%) and male (76%) with an average age of 42.2 years. Cocaine and ethanol were more frequently detected in black decedents, whereas opiates and polydrug abuse were more common in white decedents throughout the period studied. Preliminary investigation showed a high index of suspicion for the specific drug involved in virtually all opiate and alcohol cases, and in 62% of cocaine-related cases. Overall, the 100 accidental drug deaths in 2002 accounted for 7.5% of all deaths investigated and certified by the Fulton County Medical Examiner's Office. Our study provides further evidence to support the lack of correlation between serum drug levels and the mechanism of drug toxicity in cocaine-related deaths. No statistically significant differences were shown in parent cocaine or benzoylecgonine concentrations between those cases certified as toxicities or poisonings versus those cases certified as aggravating underlying disease or causing an adverse event. In addition, 62% of the cocaine-related death cases were considered initially to be of high suspicion for drug-related death, thus emphasizing the strong importance of scene information/investigative reports in evaluating drug-death cases and in formulating plans of action to handle each individual case. Among the drug-death cases handled by 6 staff medical examiners at the Fulton County Medical Examiner's Office, variation existed in autopsy performance and death certification practices. These issues are discussed in the context of the National Association of Medical Examiners' (NAME) Position Paper on Cocaine, NAME Forensic Autopsy Performance Standards, and other relevant literature. Most variations relate to completeness of the cause-of-death statement (whether or not comorbid conditions are included) rather than classification of manner of death within the office. However, specific wording in the cause of death may have significant ramifications regarding drug-related mortality statistics processed by the vital statistics system, with possible under-representation of drug-related deaths in single-cause mortality data.