Two experiments investigated relationships between state anxiety, memory processes, and children's performance on problem-solving tasks. Participants were second and sixth graders in a private elementary school in Redlands, California. In both experiments, subjects responded to three training and eight test problems presented in the introtact format. In this format, sets of cards are shown to children, one card at a time. On each card is a pair of orthogonal figures consisting of unique combinations of four dimensions (size, shape, pattern, color); dimensions have two values each. According to this method, children were shown a card and asked to point to the figure they believed contained the experimenter-defined answer for that set of cards. In both experiments, a memory aid was made available on half the test problems. Aids consisted of visual exemplars of each dimensional value and the name of that value, which was printed in block letters on the face of the card. The experiments differed only in the manner in which feedback and introtacts were delivered. In the first experiment, the stimulus card remained visible during feedback and was changed immediately after the child verbalized a hypothesis about the solution. In the second experiment, children received feedback and were required to formulate hypotheses in the absence of the stimulus cards. Subsequently, subjects were categorized into low, moderate, and high state-anxious groups on the basis of responses to questions about how they felt while solving the problems. Analysis of problem-solving behavior focused on the number of problems solved, logical rule use, and strategy use. Results indicated that memory assistance brought about more rapid solutions for both age groups; relatively little performance deficit resulted from high state-anxiety. (RH)