55 right-handed children with no family history of left-handedness received two dichotic presentations of environmental sounds. The mean test-retest interval was 9 days. Raw accuracy scores yielded acceptable temporal stability (rs > .72). The expected consistent left-ear advantage for environmental sound stimuli was only evident in younger and less mature children. Five- and 6-yr.-old children, particularly those with strong right-hand preferences, had a right-ear advantage for the stimuli. Their computed laterality coefficients showed low and nonsignificant test-retest reliabilities, however. Three- and 4-yr.-old children, particularly those with weak hand preferences, had a left-ear advantage for the same stimuli. Of the 3- and 4-yr.-old children who did not have strong right-hand preferences, 92% demonstrated consistent ear advantages across testing sessions and their laterality coefficient test-retest correlations were significant. In contrast, only 47% of the strongly right-handed 3- and 4-yr.-old children and 71% of the strongly right-handed 5- and 6-yr.-old children had consistent ear advantages for the same stimuli. It is suggested that the 3- and 4-yr.-old children processed these stimuli according to endogenous, stimulus-specific brain mechanisms and that learned processing strategies overrode these mechanisms for the 5- and 6-yr.-old children.