Human identification and forensic criminal casework may involve DNA profiling of decomposed material. Somatic microsatellite (STR) instability may lead to false exclusions and theoretically to false inclusions, both in criminal cases and in human identification. Hence, the somatic and postmortal stability of the actual sequences is crucial to the reliability of such analyses. Somatic STR stability in human tissues has been documented in small series only and the effect of postmortal tissue decomposition on microsatellite stability remains to be elucidated. On this basis, we have systematically searched for somatic STR mutations in 26 deceased humans without signs of decomposition at autopsy and 25 autopsy cases with obvious signs of postmortal decomposition. A blood sample and six tissue samples were collected from each case. Seven STRs were chosen for study, the tetranucleotides HUMVWA31/A, HUMTH01, HUMF13A1, and HUMFES/FPS, and the hyperpolymorphic markers HUMAPOAI1, D11S554 and HUMACTBP2. Denaturing gel electrophoresis was performed on an ABD Prism™ 377 gene sequencer with Genescan 672 software (Applied Biosystems, Inc.). The bone DNA profile of each case was chosen as the standard DNA profile. All cases gave profiles from additional tissues. By intraindividual comparison of DNA profiles in the cases without signs of degradation we find that the short repetitive sequences under study are stable, that is without evidence of somatic mutations. The cases with varying degree of decomposition display postmortal microsatellite stability, we detect no somatic mutations or other possible postmortal changes that could lead to between-organ non-matches. In conclusion, PCR-based STR analyses are suitable in human identification and forensic casework dealing with different tissues, even when the substrate is heavily decomposed.