The concept of primitive neuroectodermal tumour (PNET) has been evolving for many years, as has its nomenclature. It was first described as a tumour arising in peripheral nerve, when it was called neuroepithelioma [1]. These tumours are part of the differential diagnosis of malignant small round cell tumours, which include Ewing's sarcoma, rhabdomyosarcoma, neuroblastoma and lymphoma, and which appear as sheets of monotonous small round cells on light microscopy, staining dark blue with haematoxylin and eosin. In the 1970's, reports of tumours displaying neural features introduced new terms such as primitive neuroectodermal tumour and peripheral neuroectodermal tumour. A relationship to the undifferentiated primitive neuroectodermal tumours of the central nervous system in children was assumed by the title of these tumours, although it was recognised that they were not necessarily related to peripheral nerves. In 1979, Askin described a series of patients with malignant small cell tumours of the chest wall which showed neural differentiation, but which bore a resemblance to Ewing's tumour [2]. Later, this tumour was also diagnosed at other sites, and became identified with the previously described neuroepitheliomas, but the names multiplied, e.g. peripheral neuroepithelioma, peripheral neuroblastoma, Askin tumour, and peripheral PNET. The term 'primitive neuroectodermal tumour' is used in this article. Cytogenetic studies have confirmed that it forms part of a spectrum with both classical and atypical Ewing's tumour [3]. The clinical and pathological features of PNET, its management, and perspectives for the future, are discussed with reference to a case of PNET of the chest wall.