Artificial nest sites can support populations of endangered species when they are correctly installed. Here we analysed the characteristics and conditions that determined whether the northern house martin Delichon urbicum occupied more than 300 artificial nests around the city of Münster, Germany. We found that artificial nest occupation rates were influenced by various environmental and temporal factors. Positive influences included a longer time since installation and, to a lesser extent, the number of artificial nests at the same site. Negative impacts were observed from higher proportions of sealed surface cover in the surrounding area and, to a lesser extent, southward exposure. The distance to the nearest water body and the number of occupied natural nests showed no significant effect. We compared our results with descriptive evidence from the grey literature and published reports, and we give recommendations for installing artificial house martin nests for conservation practitioners. Future studies should also investigate the potential negative effects of 'dirt boards' below the nests and of gaps between the roof and the artificial nests.