Several studies have found that island populations of passerines exhibit lower levels of extrapair paternity (EPP) than mainland populations. An explanation proposing that lower levels of genetic diversity in isolated populations reduce the indirect genetic benefits of EPP to females has been supported by observational study but not tested experimentally. Here, we present the results of a manipulative study on an island population of house sparrows (Passer domesticus), which previously exhibited a significantly lower frequency of EPP than mainland populations. Fifty adults from a mainland population with significantly more genetic diversity than the island population (across 17 microsatellite loci) were introduced into the island population, and the incidence of EPP was subsequently monitored over 3 breeding seasons. In the year of the introduction, the incidence of EPP rose to approximately the level seen in mainland populations of house sparrows but dropped to an intermediate frequency in the following 2 years. Unexpectedly, parentage assignment showed that in the year of the introduction, all females producing extrapair offspring were native island birds, as were all the extrapair and cuckolded males. These results suggest that EPP in this experimental population was not driven by females trying to maximize the genetic diversity of their offspring.