The purpose of this study was to investigate the diagnostic accuracy of ultrasound (US), computed tomography (CT) and endoscopic retrograde cholangiopancreatography (ERCP) in the distinction between extrahepatic and intrahepatic causes of unjaundiced cholestasis. Moreover, attention was paid to the ability of these methods to distinguish between benign and malignant causes of obstruction. The inclusion criteria of laboratory values suggesting cholestasis were serum alkaline phosphatase 350 U/l and gammaglutamyl transpeptidase 100 U/l or liver-specific alkaline phosphatase elevated. The diagnostic value of the imaging methods was tested in 33 patients. The most common cause of unjaundiced obstruction was choledochal stone disease. The diagnostic sensitivities of US, CT and ERCP for the detection of extrahepatic cholestasis were 53%, 53% and 79%, respectively. In patients on whom all three imaging studies were done (n = 20), the difference between US and ERCP was statistically significant (p > 0.05). The specificities of US, CT and ERCP were 100%, 86% and 90%, respectively. CT was better than US in defining the benign nature of obstruction, but ERCP was superior in this respect. In malignant obstructions CT was comparable to ERCP. In conclusion, ERCP seems to be considerably sensitive for the detection of extrahepatic obstruction in cases of unjaundiced cholestasis, whereas US and CT are rather insensitive. The specificity of all these imaging methods is high. The results are in favour of CT as the best noninvasive investigation method, and also indicate the important role of ERCP. However, in many cases all three imaging methods are needed, and a flexible and complementary use of them is recommended.