Between January and July, 2003, a single blinded matched trial was done to assess focalized biliary ultrasound (US) and MR cholangiography (MRCP) compared with ERCP or surgery as reference standard in patients with suspected biliary obstruction. Focalized US was performed 24 hrs before/after MRCP. Patients were triage according symptoms, signs and laboratory in high, intermediate and low risk for obstruction. Radiologists sonographers were divided in low and highly experienced. None of the radiologist who reported either US or MRCP was aware to the other test results. Sensitivity, Specificity, PPV and NPV, (LR) Likelihood Ratios and pre and post-test odds for positive and negative test were estimated.We analized 64 out of 76 sampled patients, the prevalence of obstruction was 37 % (24 out of 64 patients). US found biliary tree dilated in 50% and MRCP in 48% of the patients. Main cause of obstruction was common bile duct stones in (16) 25%. and tumors were the second more common cause. US shown a Sensitivity (S) 58.3% when a non experienced radiologist perform the exam, and 70% when an experienced radiologist doest it. Specificity (Sp) 82,5%, PPV 66.7% and NPV 76.7%. RMCP Shown S 95.8%, Sp 87.5%, PPV 82.1%, NPV 97.2% For common bile duct stones US shown a PPV 53.3% and NPV 83.6% on the other hand MRCP a PPV 70% and NPV 95.4%. MRCP perform better than focalized US in overall patients with high risk of obstruction (37.5%), detecting the cause of obstruction but using aditional sequences rising the examination time and costs. With a positive ultrasound and LR(+) 3.3 pretest odds rise from 0.37/37% to a post-test odds of 0.55/55% and with negative test result and LR(-) 0.49 pre-test odds 0.37/37% will decrease to post-test odds of 0.2/20%. With MRCP a positive test with LR(+) 7.7 initial pre-test odds 0.37/37.5% to post-test odds of 0.8/ 80% and with a negative test result LR(-) 0.05 pretest odds 0.37/37% will decrease to post-test odds of 0.03%.Ultrasound performed by experienced sonographers has a Specificity and NPV over 80%, and perform similar to MRCP for detecting biliary tree dilatation the best predictor of obstruction. In patients with low pre-test odds (low risk) of obstruction, to use focalized US and then MRCP, when the cause of obstruction is not diagnosed could be the more cost/ efective initial alternative in this patients. This data must be confirmed with further studies with cost/effectivenes analysis.