In this study, the diagnostic yield of ultrasonography (US) in the early phase of acute urinary obstruction was retrospectively assessed and compared with that of clinical examination. 351 patients were admitted to our emergency department because of suspected renal colic over an 11 months' period: urinary obstruction was subsequently confirmed with other examinations in 76 of them, who were all submitted to renal US within 2 hours of the onset of symptoms. Emergency US results were then retrospectively compared with clinical and laboratory data and the cost of each US exam was calculated. Thirty-nine of 76 patients had negative US findings (51.3%). In the extant 37 patients, US showed hydronephrosis (which was mild in 13 patients and moderate in 6) and renal stones (in 18 patients). Hydronephrosis was associated with renal stones or increased parenchymal echogenicity in 5 and 6 patients, respectively. Pain, which was always severe, was not a useful sign for diagnostic purposes. In contrast, all the 37 patients with abnormal US findings (hydronephrosis, renal stones and increased parenchymal echogenicity) exhibited specific clinical and laboratory abnormalities such as hematuria, ketonuria and marked blood pressure increase (diastolic pressure > 100 mmHg). The same abnormalities, although less severe (mild hematuria and ketonuria and increased diastolic pressure not exceeding 100 mmHg) were nevertheless present in all the 39 patients with negative US findings and yet having a renal colic. The cost of each emergency US exam (calculated on the basis of literature tables and reference cost) was approximately It. L. 104,000. The corresponding total cost for submitting to US all the 76 patients with renal colic was approximately It. L. 7,900,000. In the early phase of urinary obstruction, the association of hematuria, ketonuria and increased blood pressure was more reliable than US findings-the latter method yielding a 51.3% false-negative rate. The presence and severity of the above clinical and laboratory abnormalities might thus represent a criterion to select the patients who are less likely to have positive US findings. This should allow the cost of medical and non-medical staff, equipment and materials to be reduced. As for our personal series, It. L. 28,000,000 was the estimated additional cost of performing US routinely in all the remaining (275) patients with suspected renal colic.