In an introductory programming course, the differential effects on learning outcomes were studied for an experimental instructional strategy that emphasized the modification and extension of existing programs (completion strategy) and a traditional strategy that emphasized the design and coding of new programs (generation strategy). Two matched groups of twenty-eight and twenty-nine high school students from grades ten through twelve volunteered for participation in a ten-lesson programming course using a small subset of the structured programming language COMAL-80. After the course, the completion group was superior to the generation group in measures concerning the construction of programs; furthermore, it was characterized by a lower mortality. The data indicated that the completion strategy facilitated the use of templates; however, this does not necessarily seem to imply that the students actually understood the working of those templates, because no differences occurred in the ability to interpret programs. In the conclusion, the completion strategy is considered to be a good alternative to more traditional strategies and recommendations are made for further improvements.