For gene therapy of the liver, in vivo applications based on adeno-associated virus are the most advanced vectors despite limitations, including low efficacy and episomal loss, potential integration and safety issues, and high production costs. Alternative vectors and/or delivery routes are of high interest. The regenerative ability of the liver bears the potential for ex vivo therapy using liver cell transplantation for disease correction if provided with a selective advantage to expand and replace the existing cell mass. Here we present such treatment of a mouse model of human phenylketonuria (PKU). Primary hepatocytes from wild-type mice were gene modified in vitro (with a lentiviral vector) that carries a gene editing system (CRISPR) to inhibit Cypor. Cypor inactivation confers paracetamol (or acetaminophen) resistance to hepatocytes and thus a growth advantage to eliminate the pre-existing liver cells upon grafting (via the spleen) and exposure to repeated treatment with paracetamol. Grafting Cypor-inactivated wild-type hepatocytes into inbred young adult enu2 (PKU) mice, followed by selective expansion by paracetamol dosing, resulted in replacing up to 5% of cell mass, normalization of blood phenylalanine, and permanent correction of PKU. Hepatocyte transplantation offers thus an armamentarium of novel therapy options for genetic liver defects.