DNA damage can cause (and result from) oxidative stress and mitochondrial impairment, both of which are implicated in the pathogenesis of Parkinson's disease (PD). We therefore examined the role of mitochondrial DNA (mtDNA) damage in human postmortem brain tissue and in in vivo and in vitro models of PD, using a newly adapted histochemical assay for abasic sites and a quantitative polymerase chain reaction (QPCR)-based assay. We identified the molecular identity of mtDNA damage to be apurinic/apyrimidinic (abasic) sites in substantia nigra dopamine neurons, but not in cortical neurons from postmortem PD specimens. To model the systemic mitochondrial impairment of PD, rats were exposed to the pesticide rotenone. After rotenone treatment that does not cause neurodegeneration, abasic sites were visualized in nigral neurons, but not in cortex. Using a QPCR-based assay, a single rotenone dose induced mtDNA damage in midbrain neurons, but not in cortical neurons; similar results were obtained in vitro in cultured neurons. Importantly, these results indicate that mtDNA damage is detectable prior to any signs of degeneration - and is produced selectively in midbrain neurons under conditions of mitochondrial impairment. The selective vulnerability of midbrain neurons to mtDNA damage was not due to differential effects of rotenone on complex I since rotenone suppressed respiration equally in midbrain and cortical neurons. However, in response to complex I inhibition, midbrain neurons produced more mitochondrial H2O2 than cortical neurons. We report selective mtDNA damage as a molecular marker of vulnerable nigral neurons in PD and suggest that this may result from intrinsic differences in how these neurons respond to complex I defects. Further, the persistence of abasic sites suggests an ineffective base excision repair response in PD.