Four major melanoma subtypes are commonly recognized: lentigo maligna melanoma (LMM), superficial spreading melanoma (SSM), acral lentiginous melanoma (ALM) and nodular melanoma (NM).1 In his unifying concept, Ackerman argues that no clinical, histological and biological criteria can been established for LMM, SSM, ALM or NM.2 The Author concludes that the best histological diagnosis should be simply “melanoma”, without referencing the artificially designated histotype, while noting the anatomic site.2 Primary melanoma evolves through logical tumor progression phases (Fig. 1).3 At first, transformed melanocytes proliferate above the epidermal basement membrane with a lentiginous or pagetoid pattern, the so-called intra-epidermal radial growth phase (RGP), corresponding to in situ LMM, SSM or ALM.4 The subsequent step is dermal invasion in the absence of tumorigenic nodules or papules and regression (non-tumorigenic micro-invasive RGP without regression).5 This phase may be followed by a tumorigenic vertical growth phase (VGP) with deeper extension into the dermis or subcutis and metastatic capacity (Fig. 2).6 An exception is represented by NM, which either skips the RGP or where the RGP is rapidly overrun by VGP.7 Besides clear-cut Ms, there are also melanocytic neoplasms with uncertain biological behaviors, which include new clinico-pathological entities more and more adopted in dermatopathology like Superficial Atypical Melanocytic Proliferation of Uncertain Significance (SAMPUS), MELanocytic Tumor of Uncertain Malignant Potential (MELTUMP), micro-invasive RGP (≤1 mm) with regression (0.75-1.00 mm in depth; >75% in volume) of uncertain tumorigenic potential at diagnosis.891011 A final diagnosis of SAMPUS or MELTUMP should always be preceded by an accurate description of the lesion with an attempt at morphological labeling, such as atypical Spitz nevus, atypical cellular blue nevus, deep penetrating nevus, pigmented epithelioid melanocytoma, dysplastic nevus with atypia or any nevus with atypical features.8 On the basis of the tumor progression, we therefore propose a new classification for the non-benignant melanocytic lesions (Table 1). It fits well with the American Joint Committee on Cancer (AJCC) staging system,12 and it should be used in all epidemiological, clinical and histological studies on M. This classification allows for explanation of the prognostic variability recorded in thin melanoma, and to identify the patients as candidates for sentinel node biopsy. In fact, in our experience, the thin melanomas with the worst prognosis are the early (≤1 mm) invasive VGP and the micro-invasive RGP with regression,6 where the regression area most probably incorporated a VGP clone.13 The sentinel node biopsy should be reserved to these two above-mentioned cases of thin melanoma, thick melanoma, and MELTUMP.