This research examined accuracy in an auditory time-to-arrival task when pe
rformance feedback was first provided to listeners and was subsequently wit
hdrawn. Listeners made judgments about the time to arrival of an approachin
g car based on various portions of the event. Listeners participated in 3 e
xperimental sessions on consecutive days. The experimental group received n
o feedback during the first session, feedback during the second session, an
d no feedback during the final session. When feedback was withdrawn, the hi
gher performance level attained during training was retained. We discuss re
sults in terms of theories of time to-arrival perception and the importance
of stimulus naturalness.