Accurate pathology reports set the foundation for successful cancer care. However, many pathologists may be confronted with a specific rare cancer perhaps once or twice in their entire professional career. This is why diagnosing rare cancers accurately can present a real challenge for a pathologist. At the same time, it is extremely important, especially in the field of rare cancers, to combine information from molecular biology, pathology and clinical practice to set up an appropriate treatment plan. The diagnostic approach to rare cancers does not differ necessarily to that for frequent cancers. Classification is still based on morphologic observation and the use of a standardized nomenclature is strongly advised. WHO “blue books” currently represent a broadly accepted classificative framework and is currently regarded as the “gold standard” for histologic classification of tumors. As mentioned microscopic observation need to be almost always implemented with immunophenotypic findings as well as with molecular genetics data. Ancillary techniques not only represent valuable diagnostic tools but also offer key prognostic and/or predictive information. When dealing with rare cancers the major challenge is obviously represented by the low frequency that, outside referral centers, may generate lack of a specific expertise. This problem tends unavoidably to affect both morphologic and molecular skills. As an example, while acceptable level of expertise in RAS testing in a frequent tumors such as colorectal cancer can be likely achieved also in small centers, the same level of accuracy cannot be met in the assessment of sarcoma-related translocations. Recently published data clearly indicate that centralized molecular testing for rare neoplasms offers significantly higher level of diagnostic accuracy. The creation of collaborative networks for molecular testing and more in general for the diagnosis of rare cancers may represent a valid solution in order to provide all rare cancer patients with precise diagnoses and consequent optimal therapeutic options.