Some recent experiments (D. A. Hay, 1975, Nature (London), 257, 44-46) investigating maze behavior of Drosophila melanogaster have been interpreted as evidence for associative behavior in this organism, while other, similar experiments (G. Bicker & H. C. Spatz, 1976, Nature (London), 260, 371) have been interpreted as evidence for nonassociative behavior processes. The experiments described below investigate differences in the behavior of flies run in mazes previously used by flies and left uncleaned, and of flies run in clean mazes. Apparently cues left by flies in a previous run may have a significant effect on the behavior of flies in a later run. The results suggest a need for a reevaluation of experiments earlier interpreted as associative behavior (Hay, 1975). © 1979 Academic Press, Inc.