The Alternate Uses Test was administered to 50 undergraduate males instructed to produce creative (i.e., novel and worthwhile) uses and to 55 comparable subjects simply instructed to produce as many uses as possible. All uses were rated for creativity. An index of self-assessed creative thinking ability correlated significantly more strongly (p less than .05) with the number of creative uses produced in the qualitatively-oriented condition than with the number of creative or total uses produced in the standard, quantitatively-oriented condition. The correlation between self-rated creative ability and creative uses production in the qualitatively-oriented condition remained significant (p less than .001) after indices of achievement motivation and general verbal aptitude were partialled out. The results were interpreted as demonstrating the value of coordinating informative divergent thinking test instructions with qualitative scoring criteria.