This study examined the effectiveness of a metacourse consisting of eight lessons interspersed over a semester-long beginning course in BASIC and aimed at providing mental models, problem-solving strategies, key concepts, and other heuristic structures. The experimental group consisted of 6 teachers who taught 9 classes of a total of 132 high school students; the control group consisted of 9 teachers who taught 13 classes of a total of 239 high school students. Data collection included a student questionnaire on previous experience with computers, a pre/posttest to assess general cognitive skills at the beginning of the term and possible transfer effects at the end, classroom observations, an end-of-semester test on BASIC, and homework assignments. Analysis focused on teachers' fidelity to the metacourse lessons, the impact of instruction on the students' mastery of BASIC, and the transfer of cognitive skills from programming to other domains. Results showed that teachers were faithful to the lessons and found them quite teachable. Experimental group students made significantly fewer errors on the test of BASIC and did significantly better on all major categories of problems. Evidence of transfer that was observed was limited to a particular problem similar to the programming tasks. The appendices include examples from the metacourse manual, the cognitive skills pre/posttest, the BASIC test, the classroom observation sheet, and the student questionnaire. (34 references) (Author/MES)