Uterine smooth muscle tumors (SMTs) are usually recognized on the basis of their routine morphologic features; however, their distinction from endometrial stromal tumors (ESTs), the second most common mesenchymal tumor of the uterus, is sometimes problematic. Histone deacetylases (HDACs) were originally identified as nuclear enzymes regulating histone acetylation. We have recently shown that in normal human tissues, HDAC8 is exclusively expressed in the cytoplasm of cells showing smooth muscle differentiation. In this study, we examined HDAC8 expression in SMTs and ESTs of the uterus to determine whether HDAC8 may be a useful diagnostic tool in the classification of problematic uterine mesenchymal tumors. HDAC8 immunohistochemical staining was performed in 15 leiomyomas (LMs), 9 highly cellular leiomyomas (HCLs), 8 epithelioid SMTs, 13 leiomyosarcomas (LMSs), and 17 ESTs, including 3 with sex-cord differentiation and 5 with smooth muscle differentiation. All tumors were also stained for other smooth muscle markers (desmin, h-caldesmon, smooth muscle actin [SMA], smooth muscle myosin heavy chain) and for CD10. All LMs had moderate to strong expression of all smooth muscle markers. HDAC8 was detected in 8 of 9 HCLs and in all epithelioid SMTs (8 of 8); however, it was weak in 4 epithelioid SMTs. In contrast, desmin, h-caldesmon and smooth muscle myosin were positive in only 2 of 8, 1 of 8 and 4 of 8 epithelioid SMTs, respectively. All smooth muscle markers had similar frequency of staining in LMSs; however, HDAC8 showed overall moderate intensity compared with other smooth muscle markers, which showed stronger staining. HDAC8, h-caldesmon, and smooth muscle myosin did not stain conventional areas of ESTs or ESTs with sex-cord differentiation, whereas SMA and desmin were positive in those areas in 4 of 12 and 3 of 12 ESTs, respectively. Areas of smooth muscle differentiation in ESTs were positive for HDAC8 in all cases, but they were less constantly positive for the other smooth muscle markers. CD10 was expressed in most ESTs (14 of 17), but it was also positive in 15 of 45 SMTs. In conclusion: 1) HDAC8 seems to be a specific marker of SM differentiation because conventional ESTs and ESTs with sex-cord differentiation are negative for HDAC8, whereas areas of smooth muscle differentiation in these tumors are consistently positive; 2) HDAC8 gives similar results to those obtained with desmin, h-caldesmon, and smooth muscle myosin in both LMs and LMSs, although the staining for HDAC8 in LMSs tends to be less intense; 3) HDAC8 may be a more sensitive marker than desmin and h-caldesmon in epithelioid SMTs. Thus, HDAC8 detection may be useful in helping the differential diagnosis of uterine mesenchymal tumors.