Homologous recombination (HR) is the process by which a double-strand break in DNA is repaired using an identical donor template. Despite rapid progress in identifying the functions of the proteins that mediate HR, little is known about how broken DNA finds its homologous template. This process, coined homology search, has been difficult to monitor experimentally. Therefore, we present here a computational approach to model the effect of subnuclear positioning and chromatin dynamics on homology search. We found that, in our model, homology search occurs more efficiently if both the cut site and its template are at the nuclear periphery, whereas restricting the movement of the template or the break alone to the periphery markedly increases the time of the search. Immobilization of either component at any position slows down the search. Based on these results, we propose a new model for homology search, the facilitated random search model, which predicts that the search is random, but that nuclear organization and dynamics strongly influence its speed and efficiency.