In the field, superparasitism of Lymantria dispar (L.) by Parasetigena silvestris (Robineau-Desvoidy) was not the result of random oviposition, but, because parasitoid eggs were aggregated, certain hosts were more likely to be parasitized than average. Parasitoid eggs were more aggregated when gypsy moth larvae were collected from under burlap bands than when larvae were collected elsewhere in the same 9-ha plot, resulting in lowered mortality due to parasitism. This finding suggests that collecting larvae from burlap bands may not provide accurate estimates of the impact of P. silvestris on populations of L. dispar. In laboratory studies, deposition of more than one egg on a single host significantly increased parasitoid emergence and host mortality. However, increasing superparasitism had a negative effect on both the probability that an individual parasitoid would survive to emerge from a host and the size of the puparium produced by the parasitoid. The probability of parasitoid survival was higher when fifth- rather than fourth-instar gypsy moth larvae were attacked, but puparia produced by parasitoids emerging from fifth-instar larvae were smaller.