In 1991, media reports of an increase in the number of deaths attributed to methadone toxicity in Harris County, Texas, raised public concern about the safety of methadone. This concern was heightened by publicity surrounding the closure of three Harris County methadone maintenance treatment programs due to their poor compliance with federal methadone regulations. In response to this concern, the Texas Department of Public Health requested that the Centers for Disease Control and Prevention (CDC) assist in an epidemiologic study to determine the extent of methadone-related mortality in Harris County during 1991 and to determine the role of methadone maintenance treatment in these deaths. We reviewed cases investigated by the Harris County Medical Examiner's Office from 1987 through 1992 in which methadone was detected by postmortem drug testing. The autopsy reports for cases occurring in 1991 were also reviewed by three independent forensic pathologists who were asked to determine the role of methadone in the death. In addition, we attempted to document Harris County methadone maintenance treatment program enrollment for each decedent. We identified 91 decedents in whom methadone was detected at the time of death, with the largest number of cases occurring in 1991 (n = 27). Other substances, including alcohol, were detected in 85% of the cases. The Harris County Medical Examiner attributed 11 of the deaths to methadone toxicity. No more than three cases per year from 1987 through 1992 were attributed to methadone toxicity. In contrast, 34 deaths were attributed to polydrug toxicity, the largest number occurring in 1991 (n = 11). There was good agreement between the results of the independent review and the opinions of the Harris County Medical Examiner. Only 20% of the decedents were found to have been enrolled in a Harris County methadone maintenance treatment program at the time of death. Four people died of drug toxicity shortly after enrolling in a methadone maintenance treatment program. We found an increase in the number deaths occurring in Harris County, Texas, in 1991 in which methadone was detected. We also found that methadone blood levels were higher among decedents identified for 1991 and 1992 than among those identified in the previous years studied. However, we did not find evidence that the cause of these deaths could be attributed solely to methadone toxicity. Instead, for all years studied, the use of multiple drugs was the leading cause of death among people in whom methadone was detected. This finding points out the difficulties involved in determining the role of methadone as a cause of death.