Despite mathematicians valuing the ability to visualize a problem and psychologists finding positive correlations of visual-spatial ability with success in mathematics, many educationists remain unconvinced about the benefits of visualization for mathematical understanding. This paper describes research that compared a "visual" to a "verbal" teaching approach through teaching a range of early secondary school mathematics topics to two classes using one or other approach. The two classes were compared through a post-intervention test of mathematical competency, on which the verbally taught class scored significantly higher. No interactions were found between teaching style and the learner's preferred style although the pupils identified as "visualizers" did tend to perform more poorly. [For complete proceedings, see ED489597.]