Introductions of biological control agents may cause bottlenecks in population size despite efforts to avoid them. We examined the population genetics of Aphidius ervi (Hymenoptera: Braconidae), a parasitoid that was introduced to North America from Western Europe in 1959 to control pea aphids. To explore the phylogeographical relationships of A. ervi we sequenced 1249 bp of mitochondrial DNA (mtDNA) from 27 individuals from the native range and 51 individuals from the introduced range. Most individuals from Western Europe, the Middle East and North America shared one of two common haplotypes, consistent with the known history of the introduction. However, some A. ervi from the Pacific Northwest have a haplotype that is most similar to haplotypes found in Japan, raising the possibility of a second accidental introduction. To examine population structure and assess whether a bottleneck occurred upon introduction to North America, we assayed variation at 5 microsatellite loci in 62 individuals from 2 native populations and 230 individuals from 6 introduced populations. Introduced samples had fewer rare alleles than native samples (F1,34 = 13.5, P = 0.0008), but heterozygosity did not differ significantly. These results suggest that a mild bottleneck occurred in spite of the introduction of over 1000 individuals. Using a hierarchical Bayesian approach, the founding population size was estimated to be 245 individuals. amova showed significant genetic differentiation between the European and North American samples, and a Bayesian assignment approach clustered individuals into four groups, with most European individuals in one group and most North American individuals in the other three. These results highlight that genetic changes are associated with founder events in rapidly growing natural populations, even when the founding population size is relatively large.