Recombination rates vary within and between species and this variation is known to be influenced by genetic and epigenetic factors. Moreover, in most eukaryotic species examined so far, crossovers occur non‐randomly along chromosomes. This variation is predicted to impact rates of evolution and levels of diversity across genomes. Here, I present an analysis of this dual property of variable recombination rates, investigated in terms of genetic and epigenetic causes and evolutionary consequences applied to the model system Drosophila melanogaster. In terms of evolutionary consequences, previous studies showed that background selection (BGS) plays a major role explaining levels of diversity across the genome, and thus BGS predictions are adequate as baseline to infer instances of balancing selection or recent selective sweeps. I now show that differences in recombination landscapes among populations of D. melanogaster, and the corresponding population-specific differences in predicted BGS, play a significant role explaining highly-localized population-specific differences in nucleotide diversity without requiring invoking local adaptation and recent selective sweeps. In terms of causes, I report analyses to investigate whether DNA motif distribution across the D. melanogaster genome can be used to predict some of the observed variation in crossover rates. This study exposes a combinatorial influence of motif presence able to account for more than 40% of the variance in crossover rates across the whole genome, an unprecedented result in any species. This high predictive power is maintained after removing sub-telomeric and -centromeric regions known to have strongly reduced crossover rates. The study also shows that transcriptional activity during early meiosis and differences in motif use between autosomes and the X chromosome add to the predictive power of the models.