The ideal classification of basal cell carcinoma (BCC) should be able to identify subtypes which correlate with clinical behaviour and treatment requirements. Unfortunately, however, such a classification has yet to be defined. In the interim, the currently most favoured classification is one based predominantly on histological growth pattern. This classification contributes to the useful concept of low- and high-risk histological subtypes of BCC. The latter are characterized by an increased probability of subclinical extension and/or incomplete excision and/or aggressive local invasive behaviour and/or local recurrence. The Royal College of Pathologists has published a minimum dataset for the histopathological reporting of BCC and this has been written to be compatible with the British Association of Dermatologists' management guidelines. Growth patterns to be reported include nodular, superficial, infiltrative/morphoeic and micronodular types, together with differentiation when of severely atypical or malignant squamous type (basosquamous carcinoma). Deep and peripheral excision margins will be reported to be either involved or clear. The latter will include a comment of a clearance of less than 1 mm for close margins and a measured distance in whole millimetres for other excisions. Clinical assessment and histology remain the 'gold standard' for evaluating BCC and cancers in general. However, in the postgenomic era emphasis is changing from the gathering and archiving of genomic data to its analysis and use in guiding clinical practice. In this context, a current goal is to define cancer phenotype in terms of molecular abnormalities and use this as a new gold standard. One way to assess whether this goal is being achieved for BCC is to determine whether our knowledge of its molecular pathology has any relevance to the minimum dataset for histological reporting. Knowledge of BCC molecular pathology has been fuelled by the recent discovery that deregulation of the Hedgehog (Hh) signalling pathway, a key player in embryonic patterning, appears to be fundamental to tumour growth. But despite accrual of a large amount of data concerning Hh pathway molecular alterations in neoplasia, little is known about the functional consequences of these changes in BCC, how they lead to tumour development, or how they relate to non-Hh pathway alterations such as TP53 mutation. Recent work suggests that the cellular localization of beta-catenin gives a degree of credence to the growth pattern classification of BCC. Furthermore, it is possible that beta-catenin may have a pathogenetic role in the invasive behaviour of BCC. This review draws on current evidence to discuss these issues and assess whether they are relevant to the minimum dataset.