The history of a 6-year-old girl with a tumor originating from thoracic spine and finally becoming resistant to surgery, radio-, and chemotherapy is reported. Tumor-biopsy material was studied by light and electron microscopy, in cell culture, by acetylcholinesterase ultracytochemistry, and by quantitative catecholamine analysis and this led to the rejection of the initial diagnosis of a neuroblastoma. Light microscopy revealed a uniform population of undifferentiated cells incompletely lobulated by broad fibrovascular septa. Using the electron microscope, cells were characterized by large intracellular pools of glycogen, little cytoplasm with an abundance of free ribosomes and a paucity of organelles. A few cells displayed desmosome-like attachment sites. Staining for specific and unspecific acetylcholinesterase was negative with light and electron microscopy, as were the results of catecholamine histofluorescence using the glyoxylic acid method. The latter result was confirmed by the negative outcome of quantitative analyses of dopamine, noradrenaline, and adrenaline with high pressure liquid chromatography nd electrochemical detection in tissue samples. Tumor cells could easily be maintained in culture for up to 4 weeks. None of a variety of treatments that are known to favor expression of neuronal characteristics in neuroblastoma cells (serum withdrawal, nerve growth factor, dbcAMP, dexamethasone) induced morphological differentiation in cultured tumor cells. On the basis of the clinical history, morphology, and of our experiments with tumor cells, the diagnosis of a so-called extraskeletal Ewing's sarcoma is most likely. Our results strengthen the view that a cell biology approach may be valuable in neuroblastoma differential diagnosis.