The purpose of this report was to test the hypothesis that extra-scope transfer depends on the extent to which a statement of strategy may be viewed as a restriction of a more general strategy. Sixty-six high school students were taught a restricted statement of one of three strategies of varying generality. Twenty-two of these students served as a control group. All subjects were tested on six problems which were based on a variant of the game "NIM". The first two problems were within the scope of the most specific strategy, the second two were within the scope of more general strategies, and the last two were within the scope of the most general strategy. Results indicate that restricted rule statements may provide an adequate basis for generalization. Also, performance on the second problem of each pair was contingent on performance on the corresponding first problems indicating that what is learned may be determined by performance on single test items and used to predict performance on additional similar-scope problems. Suggestions are made for future research. (Author/FL)