We describe here a new stable isotope-mass spectrometric technique for measuring mitochondrial DNA (mtDNA) synthesis. Growing (2–4 mo old) and weight-stable (8–10 mo old) Sprague-Dawley rats were primed with²H₂O (deuterated water) to 2.0–2.5% body water enrichment, via intraperitoneal injection, and then given 4%²H₂O in drinking water for 3–11 wk. Mitochondria were isolated from cardiac and hindlimb muscle, and mtDNA was isolated and enzymatically hydrolyzed to deoxyribonucleosides. PCR confirmed the absence of nuclear DNA contamination. The isotopic enrichment of the deoxyribose moiety of deoxyadenosine was determined by GC-MS analysis, and percent new mtDNA was calculated by comparison to genomic DNA enrichments in a tissue with nearly complete turnover (bone marrow). Initial label incorporation into deoxyadenosine of mtDNA was linear, and turnover of mtDNA was observed in nongrowing adult female rats (1.1–1.3% new mtDNA per day in cardiac and skeletal muscle). Die-away curves of mtDNA after discontinuing²H₂O administration gave a similar turnover rate constant. Human subjects were also given²H₂O for up to 6 wk, and mitochondria from platelets were isolated. Incubation with DNase removed any contaminating genomic DNA; platelet mtDNA exhibited linear incorporation from ²H₂O and reached plateau values identical to those in genomic DNA from fully turned over cells (circulating monocytes). In conclusion, replication of mtDNA can be directly measured in vivo in rodents and humans without the use of radioactivity. Use of this technique may allow improved understanding of the regulation of mitochondrial biogenesis in health and disease.