The multistep process of hepatic carcinogenesis is mirrored by the morphologic classification of lesions detectable in cirrhosis, which include large regenerative nodules (LRN), low grade dysplastic nodules (LGDN), and high grade dysplastic nodules (HGDN). The latter belong to the "bordeline malignancy" category requiring accurate distinction from well-differentiated and early hepatocellular carcinoma (HCC). Nodules in cirrhosis are usually detected by non-invasive imaging techniques, which are unable to discriminate malignant from non-malignant forms, particularly in the 1-2-cm sized group. Liver biopsy is essential in providing practical diagnostic information to hepathologists in the management of cirrhotic patients with ultrasound (US)-detectable nodules. Histologic diagnosis on liver samples is based on the accurate search of a set of cyto-architectural features (e.g. cell atypia, cell crowding, trabecular thickness, microacini) and by a supplement of histochemical (Gomori staining) and immunocytochemical stainings. The latter rely upon the search of both well established and novel markers, targeted to evaluate stromal invasion (CK7/19), the vascular pattern (ASMA and CD34), or tumor markers (including HSP70 and glipican-3). Still, the diagnostic sensitivity is limited by the type and size of sampling and by its representativity of the entire lesion. Thus, the best diagnostic approach requires the integration ofclinical, morphological, and immunocytochemical information with imaging data (i.e. US pattern, perfusional pattern, helical computed tomography/magnetic resonance pattern). Molecular data are still under evaluation as to their diagnostic efficacy in this controversial field. Discrepancies have emerged recently between Eastern and Western interpretation of these lesions, particularly in the category of "borderline" nodules that are mostly labelled as early, well differentiated HCC by Eastern pathologists and as HGDN by Western pathologsts. Novel and more objective phenotypical and molecular markers are needed to discriminate within the grey area of borderline lesions that, epidemiologically, are likely distinct between Eastern and Western geographic areas. These tools might allow a better understanding of the boundaries of the process going from high grade dysplasia to in situ HCC and from the latter to microinvasive HCC and advanced HCC, for proper clinical management and optimal therapy.