Parasitization of larch casebearer, Coleophora laricella (Hiibner), by the internal larval parasite, Agathis pumila (Ratz.), commonly has been found to decrease between successive samples in the spring. The decrease has been attributed in part to mortality of A. pumila, which prolongs the casebearer's larval stage, thereby exposing it to subsequent parasitism by other species that are intrinsically superior (Smith 1929) and to secondary parasitism of A. pumila (Graham 1949;Quednau 1970;Raske and Schooley 1979; Ismail and Long 1982). Another reason may exist for this decrease in parasitization by A. pumila: A higher percentage of parasitized than unparasitized larvae may descend from the larch foliage to the understory vegetation, which is not in the normally sampled universe (Webb 1953;Moody 1977; Denton 1979; Long and Theroux 1979;Ryan 1983). Eidmann (1965) reports that casebearers are often found on the undergrowth when high populations strip the trees, but prior to this no one had reported any connection with parasitism. This movement may result in a decreased percentage of parasitized larvae remaining on the larch foliage. Evidence presented here substantiates this hypothesis for one location, but for a more general conclusion other areas should also be examined.Through dissections of overwintering larvae, parasitization of the 1983-1984 casebearer generation on larch foliage was determined to be approximately 50% at the Tollgate plot in northeastern Oregon in December and April (Table 1). Rearing and subsequent examination of cases collected 6 June showed that parasitization by A. pumila had dropped to only 32.6%. On that date some were seen on the undergrowth but were not sampled. One week later a sample of cases from the undergrowth showed parasitization by A. pumila to be 96.3%, significantly higher than previously found on the larch foliage (x2 = 480.7, p < 0.001). Because the attack period of A. pumila had been completed in the previous summer when the larvae were solely on the larch, the higher parasitization on the undergrowth could not have been due to different attack rates in the different strata. Rather, the behavior of larvae parasitized by A. pumila differed from behavior of nonparasitized lama. Parasitization by A. pumila subsequently decreased, both on larch foliage and on the undergrowth, mainly because of secondary parasitism of A. pumila larvae and pupae by Mesopolobus sp., Habrocytus sp., Tetrastichus sp., and Gelis sp. and an increase in the incidence of premature parasite mortality. Chrysocharis laricinellae (Ratz.) was present and parasitized 7.2-9.7% of the casebearers in the final samples but was not observed to parasitize A. pumila directly. Ismail and Long (1982) report that pupation sites of parasitized and unparasitized casebearers on larch foliage are frequently different. It is now clear that many parasitized larvae leave the foliage altogether. Thus, to obtain an accurate assessment of parasitization by A. pumila, samples must be taken prior to the dispersal of parasitized ...