The purpose of this study was to teach students a way to interpret paintings metaphorically. Twenty seven art education majors were taught the clustering strategy to write metaphoric interpretations of three types of paintings (slides): realistic, abstract, and non-objective. Subjects were then randomized into two groups. Prior to using the clustering strategy, the experimental group engaged in relaxed attention exercises; the control group heard a lecture about the importance of those exercises. A rating scale was developed, literal to metaphoric. Eight judges rated the written interpretations of the paintings. The correlations for interjudge reliability were highly positive. The first hypothesis was supported: the clustering strategy increases metaphoric interpretations. The second hypothesis was not supported, but by a small margin: relaxed attention exercises do not increase metaphoric interpretations uniformly across the three types of paintings. (Author)