Summary. The purpose of the investigation was to evaluate the diagnostic value of duplex ultrasonography (DU) of thecerebral arteries in the patients with cerebrovascular insufficiency performed in private offices in Zagreb town. The investigation was performed for all patients referred to digital subtraction angiography (DSA) after Du examination performed in one of private offices (in total, 12 offices) during the period of three years. During mentioned period 127 patients were examined (100 male, 67 female, age 28–79 y, mean 62 y) according to DU exams performed in private offices. Other patients, who had DU examinations in offices other than private, were not examined in present investigation. The method of investigation was pro- spective comparison between DSA and DU findings, where DSA was used as the gold standard. The investigation was per- formed by three radiologists prospectively. They did not know DU findings before performing DSA. The pathological findings of the supraaortal arteries and especially significant stenoses (70%–99%) of the extracranial part of the carotid arteries, obliteration and ulcerated plaques were examined. For the significant stenoses of the carotid arteries, which are indication for operative treatment, DU showed sensitivity of only 63%, specificity 79%, positive predictive value 60% and negative predictive value 81%. The diagnostic difference between occlusion and subtotal stenosis of the carotid artery was also very important. DSA showed 18 occluded arteries, while DU presented 20, nine of them false positive and seven false negative (sensitivity 36%). DU diagnosed ulcerated plaque in only three out of 37 patients, including two false positive and 36 false negative results. According to results presented here, DU of the carotid arteries performed in analysed practices are not reliable and the price of DU presents expense without benefit. The insufficient education of the doctors and defective regulations of the private practices are the possible reasons for such results.