Gains in logical thinking are assumed to result from learning computer programming. Although such assumptions are commonly held, there is little empirical work on cognitive changes resulting from computer programming. The present study reports results of 15 third-grade students at a private school in Dallas who learned LOGO during the school year. One-third of the students were given 1 h/week of individual computer time (separate from in-class instruction), and the remaining students received.5 h/week of individual computer time. At the end of the school year, the 1-h group did significantly better than the.5-h group on a conditional rule-learning task. Future research should compare gains from structured languages like LOGO with those of unstructured languages and should use additional assessment techniques.