Recently, it has been shown that transjugular liver biopsy (TJLB) with three passes gives comparable specimens to percutaneous liver biopsy (PLB). The aim of this study was to evaluate the adequacy of TJLB using four passes in a consecutive series of patients, and whether using a supportive cassette can prevent fragmentation. One hundred consecutive TJLBs in 92 patients (48 transplanted), always using four passes (19-G Tru-Cut), were compared to three-pass TJLBs. The four-pass TJLB specimens were randomized at a 1:1 ratio of liver cores placed in a cassette versus not. The four-pass TJLBs, compared to three-pass TJLBs, resulted in better specimens for length (>or=25 mm: 50% vs. 35%; p = 0.026) and number of complete portal tracts (CPTs) (>or=11: 40% vs. 26%; p = 0.027), without a higher complication rate. The four-pass TJLB with >or=11 CPTs had a median length of 27 mm, and 57% of them longer than 28 mm contained >or=11 CPTs. Putting the liver biopsy cores into a cassette did not improve the fragmentation rate or adequacy of the specimen (length and number of CPTs) of TJLB. We conclude that at least four passes with TJLB should be performed when liver specimens are needed for grading and staging. Using a supportive cassette did not reduce fragmentation.