The concept of effective population size (Ne) is used widely by conservation and evolutionary biologists as an indicator of the genetic state of populations, but its precision and relation to the census population size is often uncertain. Extra-pair fertilizations have the potential to bias estimates of Ne when they affect the number of breeders or their estimated reproductive success tallied from social pedigrees. We tested if the occurrence of extra-pair fertilizations influenced estimates of Ne in a resident population of song sparrows Melospiza melodia using four years of detailed behavioural and genetic data. Estimates of Ne based on social and genetic data were nearly identical and averaged c. 65% of the census population size over four years, despite that 28% of 471 independent young were sired outside of social pairs. Variance in male reproductive success also did not differ between estimates based on social and genetic data, indicating that extra-pair mating had little effect on the distribution of reproductive success in our study population. Our results show that the genetic assignment will not always be necessary to estimate Ne precisely.