Abstract Ninety-six consecutive patients underwent sialography over a 10 year period at North Tees District General Hospital; there were 68 parotid and 28 submandibular examinations. Patients were divided into two groups: group 1, patients with symptoms but no clinical signs (n = 27); group 2, patients with swelling of the salivary gland (n = 69). Results showed that sialography was of considerable value in group 1 patients demonstrating pathology in 33 per cent (9/27 patients). In patients with a suspected, but unusually situated, parotid tumour (n = 20) sialography was confirmatory in 10 patients but the technique failed in four, gave a false negative result in four and would seem to be of limited value. In the remaining patients in group 2 with diffuse glandular swelling (n = 49) the main contribution of sialography was in demonstrating sialectasia (9), duct strictures (4) and non-opaque parotid calculi (2), but even in this group of patients 35 per cent (17/49 patients) of the examinations were normal. When submandibular calculi were demonstrable on plain radiography sialography contributed no further relevant information.