Abstract Biliary carcinoma, biliary fistula with occasional gallstone ileus and primary sclerosing cholangitis (PSC) are serious diseases and present specific diagnostic and therapeutic challenges. Stenting of biliary obstruction has also involved problems, but the reports are contradictory and partly limited. The aim of the present work was to evaluate and compare various imaging modalities in biliary diseases. The study also aimed to evaluate the usefulness of metallic stents in malignant biliary obstruction. The study population consisted of 210 patients with gallbladder carcinoma, bile duct carcinoma, biliary fistula, PSC or malignant biliary obstruction and eight control patients with various hepatobiliary diseases. The imaging findings of 80 patients with gallbladder carcinoma, 58 patients with bile duct carcinoma, and 16 patients with biliary fistula were reviewed. Nine patients with PSC underwent magnetic resonance cholangiography (MRC) and magnetic resonance imaging (MRI) of the liver, ultrasonography (US) of the liver and the bile ducts and endoscopic retrograde cholangiography (ERC). Eight control patients had had MRC and MRI of the liver and ERC. The medical records and radiographs of 39 patients with malignant biliary obstruction treated with percutaneously inserted metallic stents were also analysed. The stents included 48 Wallstents and seven Memotherm stents. In cases of gallbladder carcinoma, US visualised the primary tumour in 68 % and computed tomography (CT) in 57 % of the cases examined, but both methods were insufficient for accurate staging. In bile duct carcinoma, US revealed the primary tumour in 63 % and CT in 44 % of the cases examined. Both methods were sensitive in diagnosing peripheral intrahepatic cholangiocarcinoma, but inaccurate for more distal bile duct carcinoma or abdominal spread. The infiltrating type of gallbladder carcinoma and bile duct carcinoma were difficult to detect. US and CT were sensitive in revealing bile duct obstruction. The patients with biliary fistula and gallstone ileus had undergone various examinations with pathological, but not diagnostic results, and there was often a delay to diagnosis. Imaging did not reveal any of the ten spontaneous fistulas, but CT showed one of the five cases of gallstone ileus, and Gastrografin® meal revealed the single case of Bouveret's syndrome. Fistulography or cholangiography revealed all but one of the six iatrogenic fistulas. A nonvisualised or shrunken gallbladder at US should raise a suspicion of biliary enteric fistula in an appropriate clinical setting. MRC-MRI depicted the changes of PSC correctly in nine patients (radiologist 1) and in eight patients with one false positive finding (radiologist 2) in a blinded analysis. In the segmental comparison MRC missed especially bile duct dilatations. MRC was too pessimistic in the evaluation of the predictors of poor outcome. US detected features suggestive of PSC in eight patients (radiologist 3). US was unable to indicate the predictors of poor outcome. Of the patients with metallic stents in malignant biliary obstruction, 30 % had early and 66 % late complications, including stent obstructions, which occurred in 27 % of the patients at a mean of 4.4 months. The cause was mostly tumour ingrowth or overgrowth. The 25-week and 50-week patency rates were 71 % and 42 %. The patency rates of the patients with cholangiocarcinoma were significantly the lowest. There was also a tendency towards lower patency with less dilatation of the stents, an increasing number of the stents, longer strictures and hilar strictures. Many other complications were infectious. 31 % of the patients had late reinterventions.