The aim of this study was to evaluate whether a stimulated diuresis test associated with ultrasound is able to differentiate pelvic dilation due to atonicity from pelviureteric junction obstruction (PUO). 42 patients (25 f, 17 m) with minimal or moderate renal pelvis dilation revealed by sonography were selected for the test. Three different measurements of the anteroposterior diameter (APD) of the renal pelvis were done at the renal hilus level, by using a transversal ultrasound scan: the first under spontaneous diuresis conditions, the second after hydration with 1.5 liters of water, and the third with a full bladder 15 min after intravenous injection of 20 mg frusemide. All the patients underwent pyelography. Baseline APD (bAPD) linearly correlated with the PAD both after hydration and frusemide (r = 0.89 and r = 0.84, respectively). A descriptive evaluation of the frequency distribution of the bAPD suggested the possibility that the data samples could belong to three populations with different underlying pathophysiological conditions. Correspondence analysis between bAPD distribution and PUO suggested that the best grouping of data was: group 1 (11 patients) bAPD < 13 mm, group 2 (14 patients) 13 > or = bAPD < or = 20 mm, group 3 (17 patients) bAPD > 20 mm (likelihood ratio chi 2 46.36; d.f. = 2). Standard intravenous pyelography showed an increase in pelvis size compatible with PUO in 2 patients from group 2 and in all patients from group 3.(ABSTRACT TRUNCATED AT 250 WORDS)