Despite impressive progress in prevention and therapy of premalignant and malignant dysplasia the worldwide burden of cancer is relatively unchanged. Supplementation of the therapeutic arsenal by immunotherapeutic methods would have the potential to make a significant impact. Dysplastic lesions and cancer of the cervix show strong association with human papillomaviruses (HPV), as do tumours of other mucosal epithelia like squamous cell carcinoma of the head and neck. Such tumours are distinct from most other malignancies in that they harbour foreign antigens derived from the virus. The expression of viral oncogenes is necessary to maintain the cancerous phenotype. Therefore, these antigens are unique to the tumour and very attractive targets for ‘proof of concept’ studies in the development of therapeutic vaccinesshowing the general applicability of tumour vaccination and prove the correlation of immune response and clinical response. To date numerous clinical trials have been performed with candidate vaccines predominantly for cervical cancer and its precursors. Although a naturally induced anti-HPV T cell response in patients can be shown, the success of therapeutic vaccines has so far been limited. This can probably be attributed to immunosuppression, immunoselection and immunoediting of the tumour cells and other, mostly unknown, factors of the individual contributing to the failure of autonomous clearance of the infection. Overriding this failure, reversing immunosuppression and application in early stages of the disease are the key tasks for future development of therapeutic vaccines. This review will summarize the basis and recent developments of therapeutic vaccines and discuss obstacles that hinder their success.