The emergence of multidrug resistant bacteria has become a major cause of therapeutic failure in treating infectious diseases. Multidrug resistance is frequently acquired by horizontal gene transfers, but can also arise de novo through mutations. In the latter case, recombination may be important in reducing clonal interference between selected resistance mutations that spread simultaneously within the population. Many bacteria, including important pathogens, regularly undergo recombination via natural transformation (uptake of DNA from the environment), but the role of natural transformation in the evolution of de novo multidrug resistance evolution is unclear. Our study aims at characterizing the evolutionary dynamics of de novo multidrug resistance through evolution experiments in which the emergence and spread of resistance mutations is monitored and the impact of recombination assessed. We initiated our evolution experiment with populations comprising either naturally competent or non-competent genotypes of Acinetobacter baylyi. These populations were then propagated by serial transfer for ~650 generations under sub-lethal doses of rifampicin and streptomycin antibiotic combinations. We then characterized evolved populations by phenotypic assays and whole genome sequencing. Both growth rate and competition assays demonstrated that the populations propagated under drug pressure had evolved higher fitness when tested in same environment, but there was no difference in fitness gain between competent and non-competent populations. Moreover, antibiotic susceptibility assays showed that clones that evolved in presence of drugs had become strongly resistant to rifampicin, whereas resistance to streptomycin was weaker or absent. Consistent with these findings, whole genome sequencing revealed an abundance of different rpoB mutations (indicating target alteration as a resistance mechanism). We have also identified other mutations previously reported to be associated with resistance to other antibiotics. In conclusion, we saw no evidence that recombination by transformation facilitates adaptation to antibiotics, presumably because the low number of mutations that were spreading prevented clonal interference.