The presence of 6-monoacetylmorphine (6-MAM) is often used to distinguish between heroin (diacetylmorphine) and morphine exposures. 6-MAM, however, is rapidly metabolized to morphine and may not be present in detectable quantities in blood following heroin exposure. Recent studies have shown that 6-MAM may persist in cerebrospinal fluid (CSF) and this specimen may be preferable for establishing heroin exposure. This study reports postmortem distribution of 6-MAM, unconjugated morphine, and codeine in different tissues from 25 deceased individuals. In all cases, 6-MAM was detected in vitreous humor, and in CSF in 16 of the 25 cases (64%). When 6-MAM was detected in blood (13 of 25 cases), the level of 6-MAM in vitreous humor and CSF was higher than in blood, with a mean concentration ratio of 11.3 (range: 1.7-27) for vitreous humor and 6.6 (range: 2.6-17.3) for CSF. 6-MAM was not detected in liver in any of the cases examined. Free morphine levels were highest in liver, followed by blood, CSF, and vitreous humor. The concentration ratios (mean +/- standard deviation) for free morphine in vitreous humor, CSF, and liver to that in blood were 0.36 +/- 0.18, 0.64 +/- 0.27, and 2.99 +/- 2.12, respectively. The liver/blood ratio was consistent with previously reported values for morphine in heart and femoral blood. Codeine levels following heroin overdose were consistently low relative to the morphine concentration. For blood, liver, and CSF, the ratio of codeine to morphine was essentially the same (0.06), whereas the vitreous codeine/morphine concentration ratio was slightly higher (0.19). These results characterize the distribution of heroin metabolites in postmortem tissues. Vitreous humor appears to be a useful specimen for determining 6-MAM and establishing the morphine was derived from heroin.