The diagnostic accuracy of conventional radiography and ultrasound (US) in patients with acute or chronic abdominal pain was investigated by the authors in two prospective studies to assess the capabilities of the two imaging techniques. 262 patients (135 men and 127 women, age range: 25-83 years) were entered into the study within 8 months. The patients were divided into different groups according to clinical presentation (80 patients with acute and 182 with chronic pain) and to the referring physician (a hospital clinician for 127 patients and a general practitioner for 135 patients). For each technique the presence (positive result) or the absence (negative result) of abnormal imaging features related to clinical symptoms was recorded and the agreement of conventional radiography and US findings was assessed. Statistical analysis showed similar sensitivity of conventional radiography and US in the examination of acute hospitalized patients (54.3% positive results), chronic hospitalized patients (28.1% negative results) and chronic nonhospitalized patients (47.2% negative results). The results of our prospective studies showed poor diagnostic accuracy of the two techniques and clearly evidenced the improper use of diagnostic imaging methods by referring physicians, with a technical-oriented attitude in imaging examination requests. In the authors' opinion, a clinical-based approach by referring physicians with accurate patient selection might improve diagnostic accuracy, the cost/benefit ratio and the clinical effectiveness of imaging techniques.