In humans, objectively measured cardiovascular fitness is a good predictor of cognitive performance and of the risk to suffer from dementia. This association remains to be explained mechanistically, but altered mitochondrial function has been suggested. Here we use rats selected through 11 generations for intrinsic high (HCR) or low (LCR) running capacity, randomized into high intensity treadmill exercise, or a sedentary control group. We sacrificed the animals after three months or six months of exercise. The aim of our study was to identify any divergence in mitochondrial densities or function between sedentary and trained HCR or LCR rats. At the electron microscopic level, we examined the density and size of the mitochondria in three hippocampal areas: the stratum radiatum and the lacunosum moleculare of hippocampus CA1, and the hilus of the dentate gyrus. In LCR rats, exercise caused an increase in the mitochondrial density in the hilus, but not in the other areas, at both time-points. No effect of exercise on mitochondrial density was seen in the HCR rats. As a measure of oxidative stress (and indirectly of mitochondrial function), we analyzed the densities of superoxide dismutase in mitochondria of the animals. We further performed immunoperoxidase labelling against Iba1, showing that the average size and the total area covered by microglia was decreased in rats exposed to 6 months of exercise regardless of strain. Our data suggest that intrinsic aerobic fitness is a predictor for how the brain responds to exercise. In a wider perspective, these data raise the question whether some individuals are genetically abler to benefit from exercise than others.