Today's treatment methods have less effect on the prognosis of malignant melanomas in the primary tumor stage, irrespective of their localization, than certain prognostically important factors related to the tumor and the tumor patient. In a validation study the dominating prognostic value of tumor thickness after Breslow (1975) and the patient's gender as two independent factors is confirmed. The retrospective comparative treatment studies on the effects of elective lymph node dissection (ELND) and the extent of excision on prognosis are based on the separate evaluation of male and female patients grouped according to tumor thickness classes (0.76 to 1.5, 1.51 to 3.0 and greater than 3.0 mm). The results confirm what was to be expected according to more recent views on the pathology of melanomas (Balch et al. 1987), i.e. that regionally preventive radical measures, particularly elective lymph node dissection, have a positive effect only in a limited intermediate stage of development. Thus, it was only in a small patient group of men with melanomas of the tumor thickness class 1.5 to 3.0 mm that there was, both in relation to the total number of patients (n = 123) and to the group of head and neck melanomas (n = 30), a prognostic difference of 26% or, respectively, 44% to the favor of lymph node dissection. All other male patients as well as the female patient group exhibited better survival rates after removal of the primary tumor without subsequent elective lymph node dissection. Based on own studies and the critical consideration of published treatment studies, a number of recommendations for differentiated treatment according to acknowledged prognostic criteria (stage-specific therapy!) are given.(ABSTRACT TRUNCATED AT 250 WORDS)