A series of different ultrasonographic abnormalities detected by salivary gland echography (SGE) were investigated for their discriminant power for Sjögren's syndrome (SS) in 53 patients with either primary SS (n = 27) or secondary SS (n = 26), as well as in 90 controls. Among the controls, 26 suffered from dry mouth and/or recurrent or persistent swelling of at least one parotid or submandibular gland due to other selected disorders, while 64 were healthy, asymptomatic subjects. Mild, evident or gross inhomogeneous parenchymal patterns were the only variables selected by stepwise discriminant analysis, when comparing patients to controls. However, a mild submandibular inhomogeneity did not prove useful for such a discrimination. Based on these data, a simplified evaluation and standardised quantification of salivary involvement, as detected by SGE, is proposed using an echographic score (range 0 to 6) which assigns points to the different degrees of glandular inhomogeneity. Score values above 0 showed a sensitivity of 88.8% in primary SS and of 53.8% in secondary SS, as well as a specificity of 84.6% and of 92.2% with respect to either symptomatic or healthy controls. The lower sensitivity of SGE for patients with secondary SS presumably was a result of their milder salivary involvement.