After treatment for primary clinical Stage 1 invasive cutaneous malignant melanoma, 1086 patients were followed for a minimum of 5 years from initial operation. Patient data were retrieved from the unit's melanoma registry; 96 (8.8 per cent) were treated initially by incisional biopsy, 292 (26.9 per cent) by narrow margin excision biopsy and 698 (64.3 per cent) by wide margin excision. Logistic regression analysis was performed to assess the statistical significance of the association between the various factors. The method of initial biopsy was related to maximal tumour thickness, age, and sex. Incisional biopsy rendered 38 out of 96 (40 per cent) lesions not fully assessable on current histopathological criteria, significantly higher than for the other biopsy techniques (P less than 0.0001). Incisional biopsy did not adversely affect prognosis in terms of local recurrence and mortality. Prognosis was related to tumour thickness, age and sex of the patient, and not to biopsy technique. We recommend that all suspicious lesions should be submitted to excisional rather than incisional biopsy to avoid compromising the histological assessment, given the importance of maximal tumour thickness in determining treatment and prognosis.