The aim of this thesis is to investigate the function of a number of genes involved in mammalian DNA damage repair, in particular in repair of DNA double-strand breaks (DSBs). Among a large number of different damages that can be introduced to DNA, DSBs are especially toxic. If left unrepaired, DSBs can trigger apoptosis or induce chromosomal rearrangements that can lead to carcinogenesis. Two main pathways are responsible for repair of DSBs: homologous recombination (HR) and nonhomologous end joining (NHEJ). HR is generally an error-free mechanism that restores missing information on the basis of homologous sequence obtained from sister chromatid or homologous chromosome. By contrast, NHEJ is generally error-prone. During repair by NHEJ the DNA ends can be directly ligated or short stretches of homology at the ends can be used, leading to deletions or insertions at the site of the break. A number of genes have been identified as players in DSB repair, both in prokaryotes and eukaryotes. Many of them are found in all the kingdoms of life and are similar in aspects of their sequence and function, although there are genes characteristic only for prokaryotes or eukaryotes. Many subtle differences in function also emerge from studies on HR proteins in different species. These differences might have appeared because of diverse functions repair genes have to perform in more complexed organisms, or because the initial function(s) of a gene has been distributed over multiple paralogues. Herein I concentrate on genes involved mainly in DSB repair via HR in mammalian systems. Two members of the group of HR genes are studied: Rad54 and its paralogue Rad548. Using mice and cells deficient in these genes we try to define the role of both Rad54 and Rad54B in HR and their contribution to other cellular processes. Additionally, we investigate the link between Rad54 and Snm1, a gene originally identified as being important for interstrand cross link (ICL) repair, with regard to ionizing radiation-induced DNA damage repair.