The pathogenesis of sporadic Parkinson's disease (PD) remains enigmatic. Mitochondrial complex-I defects are known to occur in the substantia nigra (SN) of PD patients and are also debated in some extracerebral tissues. Early sequencing efforts of the mitochondrial DNA (mtDNA) did not reveal specific mutations, but a long lasting discussion was devoted to the issue of randomly distributed low level point mutations, caused by oxidative stress. However, a potential functional impact remained a matter of speculation, since heteroplasmy (mutational load) at any base position analyzed, remained far below the relevant functional threshold. A clearly age-dependent increase of the ‘common mtDNA deletion’ had been demonstrated in most brain regions by several authors since 1992. However, heteroplasmy did hardly exceed 1% of total mtDNA. It became necessary to exploit PCR techniques, which were able to detect any deletion in a few microdissected dopaminergic neurons of the SN. In 2006, two groups published biochemically relevant loads of somatic mtDNA deletions in these neurons. They seem to accumulate to relevant levels in the SN dopaminergic neurons of aged individuals in general, but faster in those developing PD. It is reasonable to assume that this accumulation causes mitochondrial dysfunction of the SN, although it cannot be taken as a final proof for an early pathogenetic role of this dysfunction. Recent studies demonstrate a distribution of deletion breakpoints, which does not differ between PD, aging and classical mitochondrial disorders, suggesting a common, but yet unknown mechanism.