Creativity can be considered a function of knowledge, imagination, and evaluation and usually programed instruction is thought to be detrimental to creative behavior since the material is highly structured and the responses are usually restricted. However, this need not be so, for one seed of creativity, student control, is inherent in programed instruction. To test the possibility of teaching creative behavior via programed instruction a program on problem solving was developed and experimentally evaluated. The experiment involved six schools and 62 subjects in each of three groups: control, program alone, and program instructor-presented. A battery of psychological tests involving various aspects of creativity was used as a pre- and posttest. The results of the experiment indicated that the program instructor-presented group scored significantly higher gains than the program alone group, which in turn was significantly higher than the control group. (WH)