Organ transplantation is the last therapeutic choice for end-stage liver failure, which is limited by the lack of sufficient donors. Decellularized liver can be used as a suitable matrix for liver tissue engineering with clinical application potential. Optimizing the decellularization procedure would obtain a biological matrix with completely removed cellular components and preserved 3-dimensional structure. This study aimed to evaluate the decellularization efficacy through three anatomical routes. In this experimental study, rat liver decellularization was performed through biliary duct (BD), portal vein (PV), and hepatic vein (HV); using chemical detergents and enzymes. The decellularization efficacy was evaluated by measurement of DNA content, extracellular matrix (ECM) total proteins, and glycosaminoglycans (GAGs). ECM preservation was examined by histological and immunohistochemical (IHC) staining and scanning electron microscopy (SEM). Scaffold biocompatibility was tested by the MTT assay for HepG2 and HUVEC cell lines. Decellularization through HV and PV resulted in a transparent scaffold by complete cell removal, while the BD route produced an opaque scaffold with incomplete decellularization. H and E staining confirmed these results. Maximum DNA loss was obtained using 1% and 0.5% sodium dodecyl sulfate (SDS) in the PV and HV groups and the DNA content decreased faster in the HV group. At the final stages, the proteins excreted in the HV and PV groups were significantly less than the BD group. The GAGs level was diminished after decellularization, especially in the PV and HV groups. In the HV and PV groups the collagen amount was significantly more than the BD group. The IHC and SEM images showed that the ECM structure was preserved and cellular components were entirely removed. MTT assay showed the biocompatibility of the decellularized scaffold. The results revealed that the HV is a more suitable route for liver decellularization than the PV and BD.