Liver tropism potentially leading to massive hepatocyte transduction and hepatotoxicity still represents a major drawback to adenovirus (Ad)-based gene therapy. We previously demonstrated that substitution of the hexon hypervariable region 5 (HVR5), the most abundant capsid protein, constituted a valuable platform for efficient Ad retargeting. The use of different mouse strains revealed that HVR5 substitution also led to dramatically less adenovirus liver transduction and associated toxicity, whereas HVR5-modified Ad were still able to transduce different cell lines efficiently, including primary hepatocytes. We showed that HVR5 modification did not significantly change Ad blood clearance or liver uptake at early times. However, we were able to link the lower liver transduction to enhanced HVR5-modified Ad liver clearance and impaired use of blood factors. Most importantly, HVR5-modified vectors continued to transduce tumors in vivo as efficiently as their wild-type counterparts. Taken together, our data provide a rationale for future design of retargeted vectors with a safer profile.