For the purpose of investigating the relation between hair root morphology and STR (Short tandem repeat) typing, the root morphology of head hairs collected by sweeping, combing and pulling from twenty Japanese males was classified as follows. Hair roots swept or combed were classified into four types; A: with no root sheath, B: with little root sheath at the proximal end, C: with medial root sheath at the proximal end or little root sheath around the telogen club, and D: with much root sheath around the telogen club. Hair roots pulled were classified into three types; E: with no root sheath, F: with little root sheath, and G: with much translucent root sheath tissue around the anagen growth bulb. Histomorphological examination of hair roots showed that root sheaths were not observed around telogen clubs of type A, and root sheaths were abundantly observed around anagen growth bulbs of types F and G. In roots of type E, root sheaths were scarcely observed but nuclei were observed in the hair cortex of the anagen growth bulb. In roots of telogen or catagen phases such as types B and C, distinct root sheaths were not observed but translucent process-like structures with blue-stained nuclei were observed at the bottom part of the telogen club. When roots that adhered materials observed around the root of type D were opaque, nuclei were almost never observed. From the hair roots of eleven individuals arbitrarily selected from twenty Japanese males, STR typing was examined using the AmpFlSTR profiler kit. As a result, no STR loci were detected from all eleven individuals in roots of type A, but all STR loci were detected from them in roots of types E, F and G. In roots of types B, C and D, the STR loci of each hair showed a multifarious detection pattern. When roots adhered materials at the proximal end of the root were translucent, the possibility of STR detection was relatively high. But, when roots adhered materials not at the proximal end of the root and were opaque, the possibility of detection became low. These results suggest that the detection of STR loci from hair root samples is not influenced by the quantity but the quality of adhered materials observed in the hair roots.