Artificial nest experiments have been used in approximately 80 recent studies to identify a wide range of factors influencing reproductive success of birds. Despite the same factors being examined in different studies, few consistent patterns have emerged from these experiments. Analysis of reproductive success at artificial and natural nests reveals that reproductive success measured with use of artificial nests is frequently underestimated. In addition, several studies suggest that different species of predators may be differentially attracted to the two types of nest. These results call the utility of artificial nests into question as a tool for elucidating factors influencing the success of natural nests. Analysis of the methodology used in artificial nest experiments reveals that in general there is poor correspondence between the appearance of experimental setups and the natural systems they attempt to model. If future experiments with artificial nests are to contribute to theory, the quality of experiments must improve and the assumptions underlying the artificial approach must be tested. Collecting this information will be relatively simple compared with the time-consuming task of experimenting with natural nests. Until these data are collected, conclusions about nesting ecology derived from the use of artificial nest experiments should be treated as preliminary.