Build-up of proactive interference (PI) with visual-picture and auditory-verbal input modalities and the subsequent release from Pi following a change in modality was investigated in three experiments with boys and girls, as follows: Experiment I (n=64) at two mean age levels, 7-6 and 10-5; Experiment II (n=64) at mean age 7-6; and Experiment III (n=48) at age 11-4. PI build-up occurred in both modalities for all ages tested. Release from PI occurred following a change from auditory to visual input by not from PI occurred following a change from auditory shift. In the final experiment, this asymmetrical improvement in performance was dependent upon an interaction between the modality of the input and distractor task on the final or release trial; changing to visual input produced a release effect regardless of the distractor task modality, while auditory input was associated with improvement in recall if a visual distractor task was employed whether or not a shift in input modality had occurred. This improvement was hypothesized to represent a decrease in retroactive interference rather than a release from proactive interference.