Woolly apple aphid,Eriosoma lanigerum(Hausmann), populations were studied in eastern West Virginia from 1985 to 1989. In an unsprayed orchard, peak abundance of arboreal populations was 22–24 colonies per tree in early June each year. Spraying the orchard with a pyrethroid three times during 1989 had little effect on the population behavior, demonstrating the resilience of the woolly apple aphid and its natural enemy guild. Nearly 20% of the aphid colonies in June had syrphid larvae present and parasitism byAphelinus mali(Haldeman) was>50% in July. Age structure of arboreal woolly apple aphid colonies varied through the summer with a significant reduction in first instars in July, signalling a return of aphids to the edaphic from the arboreal environment at that time. Samples of arboreal populations were not useful for predicting year-to-year population abundance or the extent of root infestations in a managed orchard. Microhabitat preference of arboreal colonies during the spring was for wound sites and other protected feeding sites on the tree branches and trunk. Leaf axils were the predominant microhabitat (51% of the colonies observed) from the end of May through August. Cicada oviposition sites were also highly preferred, with one orchard having 98% of the colonies in cicada oviposition scars. Woolly apple aphid colonies were observed more often in wounds and protected sites on branches in sprayed orchards and in high density populations than in unsprayed or low density populations. We suggest that these protected sites act as refugia for woolly apple aphid populations in apple orchards.