Because of the poor prognosis and of oral mucosal melanoma, and patients’ short survival, large, randomised, clinical studies are difficult. We have investigated its demographic characteristics and analysed the effect of treatment, resection margins, and metastases on survival. We recorded age, sex, site of primary tumour, and types of treatment, survival, and metastases in 74 patients treated at the Department of Oral and Maxillofacial Surgery, Seoul National University Dental Hospital. Survival was analysed based on bony invasion, depth of invasion, and resection margins, and we found that it varied depending on the primary site (p=0.002), and declined with liver (p=0.001) or brain (p=0.033) metastases. The two-year survival according to the primary site was as follows: palate 85% (n=32), anterior maxillary gingiva 53% (n=13), mandible 58% (n=13), and posterior maxillary gingival 74% (n=10) and buccal mucosa 50% (n=4). The two-year survival was 34% (n=8) in patients with liver metastases and 23% (n=7) in patients with brain metastases. In cases of bony invasion (p=0.005), depth of invasion (p=0.042), unclear resection margin (p=0.023), or higher T stages (p=0.009), the survival declined considerably. Neck dissection did not affect survival (p=0.343). Survival of the patients given chemotherapy was significantly lower (p=0.013) and the two-year survival was 54.0%. The patients given radiotherapy showed no significant difference in survival compared with those not given radiotherapy (p=0.107). In conclusion, primary site, bony invasion, resection margins, depth of invasion and systemic metastases were critical to predict prognosis and selection of treatment of oral mucosal melanoma.