Most pathogenic mtDNA mutations are heteroplasmic and there is a clear correlation between high levels of mutated mtDNA in a tissue and pathology. We have found that in vivo double-strand breaks (DSBs) in mtDNA lead to digestion of cleaved mtDNA and replication of residual mtDNA. Therefore, if DSB could be targeted to mutations in mtDNA, mutant genomes could be eliminated and the wild-type mtDNA would repopulate the cells. This can be achieved by using mitochondria-targeted restriction endonucleases as a means to degrade specific mtDNA haplotypes in heteroplasmic cells or tissues. In this work, we investigated the potential of systemic delivery of mitochondria-targeted restriction endonucleases to reduce the proportion of mutant mtDNA in specific tissues. Using the asymptomatic NZB/BALB mtDNA heteroplasmic mouse as a model, we found that a mitochondria-targeted ApaLI (that cleaves BALB mtDNA at a single site and does not cleave NZB mtDNA) increased the proportion of NZB mtDNA in target tissues. This was observed in heart, using a cardiotropic adeno-associated virus type-6 (AAV6) and in liver, using the hepatotropic adenovirus type-5 (Ad5). No mtDNA depletion or loss of cytochrome c oxidase activity was observed in any of these tissues. These results show the potential of systemic delivery of viral vectors to specific organs for the therapeutic application of mitochondria-targeted restriction enzymes in mtDNA disorders.