Targeted nanomedicine holds enormous potential for advanced diagnostics and therapy. Although it is known that nanoparticles accumulate in liver in vivo, the impact of cell-targeting particles on the liver, especially in disease conditions, is largely obscure. We had previously demonstrated that peptide-conjugated nanoparticles differentially impact macrophage activation in vitro. We thus comprehensively studied the distribution of gold nanorods (AuNR) in mice in vivo and assessed their hepatotoxicity and impact on systemic and hepatic immune cells in healthy animals and experimental liver disease models. Gold nanorods were stabilized with either cetyltrimethylammonium bromide or poly(ethylene glycol) and additional bioactive tripeptides RGD or GLF. Gold nanorods mostly accumulated in liver upon systemic injection in mice, as evidenced by inductively coupled plasma mass spectrometry from different organs and by non-invasive microcomputerized tomography whole-body imaging. In liver, AuNR were only found in macrophages by seedless deposition and electron microscopy. In healthy animals, AuNR did not cause significant hepatotoxicity as evidenced by biochemical and histological analyses, even at high AuNR doses. However, flow cytometry and gene expression studies revealed that AuNR polarized hepatic macrophages, even at low doses, dependent on the respective peptide sequence, toward M1 or M2 activation. While peptide-modified AuNR did not influence liver scarring, termed fibrosis, in chronic hepatic injury models, AuNR-induced preactivation of hepatic macrophages significantly exacerbated liver damage and disease activity in experimental immune-mediated hepatitis in mice. Bioactively targeted gold nanoparticles are thus potentially harmful in clinically relevant settings of liver injury, as they can aggravate hepatitis severity.