The term "targeted cancer therapies" refers to treatment strategies designed to inhibit the product of an oncogene involved in the process of neoplastic transformation. Different categories of targeted therapies can be identified: 1) Therapies directed at oncogenes that are directly involved in the initiation of neoplastic transformation: the use of imatinib for the treatment of CML or GIST is the classical model in this subgroup. Single agent targeted therapies generally produce high response rates in this situation. 2) Therapies directed at oncogenes involved at a later stage of neoplastic transformation. These oncogenes contribute to tumor progression but not necessarily to the onset of malignant transformation. The use of trastuzumab for HER2-amplified breast adenocarcinoma is the classical model in this subgroup. These treatments are associated with low response rates when used as single agent therapy, whereas generally displaying a synergistic or additive effect with classical chemotherapy in models currently available. In contrast, when these targeted therapies are applied to tumor models where the targeted gene is present but not directly involved in the process of malignant transformation, no antitumor efficacy is generally observed. Recently, the identification of HER1 mutations in subsets of lung carcinoma as a predictive factor for response to gefitinib and erlotinib provided an example of how the empiric use of a targeted treatment may enable to identify new nosological entities. The present paper reviews examples of targeted cancer therapies and their results.