The effects of rule knowledge were investigated using Braille inkprint pairs. Both recognition and recall were studied in three groups of subjects: rule knowledge, rule discovery, and no rule. Two hypotheses were tested: (1) that the group exposed to the rule would score better than would a discovery group and a control group; and (2) that all three groups would have higher scores on the recognition task than on the recall task. Subjects included 119 undergraduate students randomly divided into three groups. Members of the rule knowledge group were told the underlying pattern in the braille alphabet, and their worksheets reflected this pattern. Members of the discovery group were not told the underlying pattern, although their worksheets allowed discovery of the rule. Members of the control group were not told about the pattern, and their worksheets were arranged to make discovery extremely unlikely. Subjects were asked to store and retrieve 26 specific dot pattern-inkprint letter pairs. The results indicate that the rule knowledge group outperformed the other two groups, but there was no difference between the recall and recognition performance. Results are discussed in relation to current models of rule learning and differences between recognition and recall procedures. Two data tables, one bar graph, and four sample worksheets are provided. (RLC)