The fact that a child knows a certain verbal relationship does not imply that this knowledge will be spontaneously employed in encoding verbal material. For example, while secondgrade children may understand connotative meaning, they might not encode rapidly presented items along a good-bad dimension. In fact, Cermak, Sagotsky, and Moshier (1972) modified Wickens' (1970) release-from-proactive-interference technique for use with children and concluded that encoding along an evaluative dimension develops after the second grade. This procedure is potentially of great value in exploring age changes in encoding.This paper is addressed to an ambiguity in Cermak et al.'s data. With sixth graders, they observed both the buildup of proactive interference over several trials with "good" or "bad" items and its release following a change in connotative meaning; these findings imply encoding along an evaluative dimension. However, fourth graders showed proactive interference but no release, and there was no buildup of interference for second graders. These results are difficult to interpret. It could be that the paradigm is sensitive to age differences in encoding, as Cermak et al. suggested, and that second graders do not encode on the basis of evaluative meaning; or, the technique may be inappropriate for work with young children. To distinguish between these possibilities, it is necessary to demonstrate that release from proactive interference can be obtained with these subjects when stimuli from a more obvious dimension are used.