We reviewed a series of deaths in which methamphetamine was detected in the decedent's blood. Analysis of postmortem whole blood was performed by gas chromatography/mass spectrometry with a limit of quantitation of 0.05 mg/L. Methamphetamine was detected in 146 cases; 52 were drug caused, i.e., a death in which the direct toxic effects of the drug caused or contributed to the death, 92 were classified as drug related, i.e., a death in which the drug was demonstrated in the blood, but did not directly cause death. A large proportion of the deaths resulted from homicidal (27%) or suicidal (15%) violence. An examination of methamphetamine concentrations in drug related deaths (n = 92), suggests that the range of concentrations in the recreational abusing population is substantial (0.05-9.30 mg/L) but with a median concentration of 0.42 mg/L, and with 90% of that population having concentrations less than 2.20 mg/L. There was substantial overlap in methamphetamine concentration between drug related deaths and drug caused deaths, although the highest concentrations were seen in the unintentional (accidental or undetermined) drug caused deaths. Methamphetamine related traffic deaths (n = 17) showed patterns of driving behavior consistent with reports elsewhere, and showed blood methamphetamine concentrations ranging from 0.05-2.60 mg/L (median 0.35 mg/L). The data show that most methamphetamine deaths occur with blood concentrations greater than 0.5 mg/L, but can occur with levels as low as 0.05 mg/L, though usually in conjunction with other drugs or significant natural disease. Neither apparently toxic nor therapeutic concentrations should be used in isolation to establish conclusively whether a death was caused by methamphetamine; proper classification of deaths involving methamphetamine requires complete death investigation, including investigation of the scene and circumstances of death, and a complete autopsy.