AIM: To present the experience and outcomes of the surgical treatment for the patients with anorectal melanoma from the Cancer Hospital, Chinese Academy of Medical Sciences. METHODS: Medical records of the diagnosis, surgery, and follow-up of 56 patients with anorectal melanoma who underwent surgery between 1975 and 2008 were retrospectively reviewed. The factors predictive for the survival rate of these patients were identified using multivariate analysis. RESULTS: The 5-year survival rate of the 56 patients with anorectal melanoma was 20%, 36 patients underwent abdominoperineal resection (APR) and 20 patients underwent wide local excision (WLE). The rates of local recurrence of the APR and WLE groups were 16.13% (5/36) and 68.75% (13/20), (P = 0.001), and the median survival time was 22 mo and 21 mo, respectively (P = 0.481). Univariate survival analysis demonstrated that the number of tumor and the depth of invasion had significant effects on the survival (P < 0.05). Multivariate analysis showed that the number of tumor [P = 0.017, 95% confidence interval (CI) = 1.273-11.075] and the depth of invasion (P = 0.015, 95% CI = 1.249-7.591) were independent prognostic factors influencing the survival rate. CONCLUSION: Complete or R0 resection is the first choice of treatment for anorectal melanoma, prognosis is poor regardless of surgical approach, and early diagnosis is the key to improved survival rate for patients with anorectal melanoma. AIM: To present the experience and outcomes of the surgical treatment for the patients with anorectal melanoma from the Cancer Hospital, Chinese Academy of Medical Sciences. METHODS: Medical records of the diagnosis, surgery, and follow-up of 56 patients with anorectal melanoma who underwent surgery between 1975 and 2008 were retrospectively reviewed. The factors predictive for the survival rate of these patients were identified using multivariate analysis. RESULTS: The 5-year survival rate of the 56 patients with anorectal melanoma was 20%, 36 patients underwent abdominoperineal resection (APR) and 20 patients underwent wide local excision (WLE). The rates of local recurrence of the APR and WLE groups were 16.13% (5/36) and 68.75% (13/20), (P = 0.001), and the median survival time was 22 mo and 21 mo, respectively (P = 0.481). Univariate survival analysis demonstrated that the number of tumor and the depth of invasion had significant effects on the survival (P <0.05). Multivariate an alysis showed that the number of tumor [P = 0.017, 95% confidence interval (CI) = 1.273-11.075] and the depth of invasion (P = 0.015, 95% CI = 1.249-7.591) were independent prognostic factors influencing the survival rate CONCLUSION: Complete or R0 resection is the first choice of treatment for anorectal melanoma, prognosis is poor regardless of surgical approach, and early diagnosis is the key to improved survival rate for patients with anorectal melanoma.