The usefulness of fine-needle aspiration cytology (FNAC) in the diagnosis and treatment of submandibular gland lesions is not well known. The 210 FNACs taken from submandibular gland lesions at Turku University Central Hospital between 1984 and 1991 were reviewed. Of these FNACs, 78 samples from primary lesions were confirmed histologically. Within this subset 10 FNACs were taken from benign neoplasms, all of which were correctly classified (sensitivity 100 per cent; specificity 88 percent). Only four of the 14 FNACs from malignant lesions were cytologically considered malignant (sensitivity 29 per cent). On the other hand, four FNACs raised a false suspicion of malignancy (specificity 6 per cent). Out of 54 FNACs from non-neoplastic lesions 43 were correct (sensitivity 80 per cent; specificity 63 per cent). There were 104 patients (123 FNACs), who had not been operated on: the follow-up of these patients shows that in this subset of FNACs there were no false malignant but probably one false benign finding (1 per cent). We conclude that FNAC can offer valuable information about the type of the submandibular gland lesion, but the decision of operative and other treatment should not be based solely on the result of FNAC.