Successful gene therapy approaches for metachromatic leukodystrophy (MLD), based either on hematopoietic stem/progenitor cells (HSPCs) or direct central nervous system (CNS) gene transfer, highlighted a requirement for high levels of arylsulfatase A (ARSA) expression to achieve correction of disease manifestations in the mouse model. Full assessment of the safety of ARSA expression above physiological levels thus represents a prerequisite for clinical translation of these approaches. Here, using lentiviral vectors (LVs), we generated two relevant models for the stringent evaluation of the consequences of ARSA overexpression in transduced cells. We first demonstrated that ARSA overexpression in human HSPCs does not affect their clonogenic and multilineage differentiation capacities in clonogenic assays and in a neonatal hematochimeric mouse model. Further, we studied ARSA overexpression in all body tissues by generating transgenic mice overexpressing the ARSA enzyme by LV up to 15-fold above the normal range and carrying multiple copies of LV in their genome. Characterization of these mice demonstrated the safety of ARSA overexpression in two main gene therapy targets, HSPCs and neurons, with maintenance of the complex functions of the hematopoietic and nervous system in the presence of supraphysiological enzyme levels. The activity of other sulfatases dependent on the same common activator, sulfatase-modifying factor-1 (SUMF1), was tested in ARSA-overexpressing HSPCs and in transgenic mice, excluding the occurrence of saturation phenomena. Overall, these data indicate that from the perspective of clinical translation, therapeutic levels of ARSA overexpression can be safely achieved. Further, they demonstrate an experimental platform for the preclinical assessment of the safety of new gene therapy approaches.