This research tested for an hypothesised interaction between instructional technique and working memory span level in outcomes from teaching how to correctly solve instances of a particular complex verbal problem form, the double modifier problem form (Casey, 1993), a problem form that arguably stresses working memory. The highest and lowest scoring sets of 22 University students (N = 64) in a working memory span test were randomly allocated to either an algebraic or a predominantly verbal instructional technique group, resulting in a 2 times 2 between groups design (n = 11 per group). It was expected that the verbal method would differentiate between low and high working memory span participants because the verbal method allowed potential distractors to be influential, a feature of problems that provide difficulties for participants with low working memory spans. In contrast, because the algebraic method avoided distractors no such effect was expected. The results supported the hypothesised interaction. The results caution that when developing teaching techniques for topics that load heavily on working memory the techniques should be tested for potential distractors that may hinder learning by people relatively low on working memory capacity.