Cutaneous malignant melanomas of the head and neck are prognostically engimatic. In addition to known prognostic determinants of stage and lesion microstage, lesion location also appears to have prognostic importance. The authors have reviewed a series of 83 microstaged head and neck melanoma patients in order to analyze the relative importance of these factors. There were 36 males and 47 females with a median age of 56 years. Eighty-one percent had pathologic Stage I disease, 7% were Stage II, and 12% were Stage III. The primary location was face in 32 patients, neck in 18, ear in 12, and scalp in 21 patients. The actuarial 5-year survival according to lesion thickness was 86% for melanoma less than 1.0 mm, 56% for 1 to 2 mm thick lesions, 47% for 2.1 to 4 mm thick lesions, and 25% for melanomas greater than 4.0 mm. The 5-year survival according to lesion location was 78% for facial and 58% for neck melanomas; for ear and scalp, the respective survivals were 33% and 37%. Median thickness was 2.0 mm for facial and 1.85 mm for neck lesions. It was 2.7 mm for ear and 2.0 mm for scalp lesions (differences not significant). There were no microstage factors that correlated with the adverse prognosis seen with scalp and ear melanomas. Multivariate analysis in the entire series (all clinical stages) showed the following to be significant: stage, thickness, and location of the primary melanoma (all less than 0.0002). In clinical Stage I melanoma, the significant prognostic factors were location (P = 0.035), thickness (P = 0.008), level (P = 0.024), and ulceration (P = 0.035). The prognosis of head and neck melanoma is uniquely influenced by location of the primary lesions in addition to stage, thickness, level, and ulceration, as observed with other cutaneous melanomas at other sites. Ear and scalp melanomas are high-risk lesions whose poor prognosis is not readily explained by any of the microstage factors reviewed.