Analogous to mixed tumors of salivary glands (" pleomorphic adenomas" ), cutaneous mixed tumors (" chondroid syringomas" ) contain a ductal (epithelial) component and a variably prominent myoepithelial component. Tumors showing purely myoepithelial differentiation (myoepitheliomas) have only recently been recognized to arise in the dermis, and to date very few cases have been described. To characterize these tumors further, 14 cutaneous myoepithelial tumors were retrieved from the authors' consult files. Eleven patients were male and 3 were female; their median age was 22.5 years (range, 10 to 63 years), and 7 patients were between 10 and 20 years old. Tumor size ranged from 0.5 to 2.5 cm (mean, 1.1 cm). Most tumors arose on the extremities: 6 on the upper limbs, 6 on the lower limbs, and 1 each on the back and nose. Ten tumors were limited to the dermis, and 5 also extended into superficial subcutis. Thirteen tumors were myoepitheliomas (lacking ductal differentiation), and 1 tumor was a myoepithelial carcinoma (exhibiting severe cytological atypia and a high mitotic rate). Histologically, 7 tumors were solid, composed of ovoid to spindled, histiocytoid, or epithelioid cells with no significant stroma, and 7 were predominantly lobulated, with cords or nests of epithelioid, plasmacytoid, or spindled cells with a variably reticular architecture and a chondromyxoid or collagenous/hyalinized stroma. One tumor was composed solely of plasmacytoid (hyaline) cells, and 1 exhibited extensive adipocytic differentiation. Among the 13 myoepitheliomas, mitoses ranged from 0 to 6 per 10 high-power fields (HPFs) (mean, 1.5); 8 tumors contained no mitoses. The myoepithelial carcinoma had 39 mitoses per 10 HPFs. By immunohistochemistry, all cases were reactive for epithelial markers (keratins and/or epithelial membrane antigen [EMA]); 13 of 14 (93%) expressed S-100 protein, 10 of 11 expressed (91%) calponin, 11 of 14 (79%) expressed EMA, 9 of 14 (64%) expressed keratins, 8 of 14 (57%) expressed smooth muscle actin, 7 of 14 (50%) expressed glial fibrillary acidic protein, 3 of 11 (27%) expressed p63, and 1 of 6 (17%) expressed desmin. All 5 cases without keratin staining were diffusely positive for EMA, and all of these cases showed a solid growth pattern. Follow-up was available for 8 patients (median follow-up, 40 months; range, 6 months to 9 years); 3 tumors (38%) recurred locally, and 1 tumor (13%) also metastasized to the lymph nodes. The case that resulted in recurrence and metastasis had the highest mitotic rate (6 per 10 HPFs) of the cytologically benign tumors. Follow-up information was not available for the myoepithelial carcinoma. This study suggests that approximately 50% of cutaneous myoepitheliomas are distinctive lesions composed of a solid proliferation of cells with abundant eosinophilic syncytial cytoplasm, which often lack immunostaining for keratin, whereas the remainder demonstrate focally reticular architecture and myxoid stroma or plasmacytoid cells, similar to their counterparts in salivary gland and soft tissue. Whereas most cutaneous myoepitheliomas behave in a benign fashion, there is apparently a significant risk for local recurrence but a low metastatic potential.