D-dimer assay was performed on 145 cerebrospinal fluid (CSF) samples from patients with or without neoplastic diseases. Levels of D-dimers were significantly higher in carcinoma and lymphoid malignancies with clinical or biological evidence of central nervous system (CNS) involvement than in diseases without such complications. In one patient, serial determinations of D-dimers were well correlated with the appearance and disappearance of CNS involvement. Although this test is not specific for neoplastic affections, our data suggest that the measurement of D-dimers in CSF may be useful in the diagnosis of CNS involvement of neoplastic cells and in monitoring intrathecal therapy in patients with lymphoma, acute lymphoblastic leukaemia or carcinoma. In this study, the D-dimer assay was also positive in some non neoplastic diseases, but failed to differentiate subarachnoid haemorrhage from traumatic lumbar puncture.