Diel drift samples utilizing nets with mesh size less than 200 microns were taken in Linesville Creek, Pennsylvania, an eastern deciduous forest stream, and Inlet Run, Wyoming, an alpine snow melt stream. Identification of drifting Chironomidae larvae to lowest level taxonomic categories indicated 51 species or species group categories representing 51.95% of the total insect drift in Linesville Creek and 18 species or species group categories representing 70.47% of the total insect drift in Inlet Run. Orthocladiinae were the predominant larvae in the drift in Linesville Creek, with 19 species comprising 43.84% of the Chironomidae drift. In decreasing abundance were Chironomini (12 species, 40.36% of Chironomidae drift), Tanytarsini (10 species, 8.89%), and Tanypodinae (10 species, 6.91%). By contrast, Diamesinae were the predominant larvae in the drift in Inlet Run, (5 species, 71.43%) followed by Orthocladiinae (10 species, 27.25%), Tanytarsini (2 species, 1.20%), and Podonominae (1 species, 0.12%). Comparison of drift composition with substrate samples and/ or emergence data indicated a close relationship between relative abundance in drift and relative abundance in the benthos. Behavioral drift patterns with nocturnal peaks were seen for 3 species or species groups in Linesville Creek. Four species with diurnal drift peaks were present in Inlet Run. Analysis of the size distribution of drifting larvae indicates that a mesh size as small as 200 microns is required to resolve diel drift patterns. It is postulated that random factors greatly influence the apparent diel drift pattern of Chironomidae when nets with mesh size in excess of 400 microns are employed in drift studies. Conflicting literature reports of behavioral drift for Chironomidae may be due to differing species composition of drifting larvae and net mesh size related artifacts.