Fifteen autopsy cases with little putrefaction were selected. Fairly good correlations were observed between blood and pericardial fluid for all drugs, neutral and basic drugs, and acidic drugs. Correlations of drug concentrations between blood and cerebrospinal fluid/femoral muscle were not as good as those between blood and pericardial fluid. No correlations were observed between blood and urine/bile. Ratios of pesticide concentrations in each specimen to those in blood showed large variation. Although the study was limited to a small number of cases, the authors determined that pericardial fluid was a good sample for quantitative confirmation of analyses performed on blood samples or as a quantitative alternative to blood in exsanguinated victims. Cerebrospinal fluid, urine, bile, and skeletal muscle were suitable only for qualitative analyses. 22 references 4 tables, and 3 figures