Two experiments are reported in which different batteries of specially designed spatial tasks were administered to male college students. The subjects were selected to be either high or relatively low in spatial visualization ability as assessed by performance on four paper-and-pencil tests (Paper Folding, Surface Development, Form Board, and Cube Comparisons). Three hypotheses proposed to account for individual differences in spatial visualization ability were investigated. These hypotheses attribute differences in spatial visualization ability to variations in: (a) representational quality, (b) transformational efficiency, and (c) preservation of representations during transformations. The failure to find differences related to spatial visualization ability in the accuracy of recognition memory decisions and in the speed of transformations is inconsistent with the first two hypotheses. The evidence was somewhat mixed with respect to the preservation-under-transformation hypothesis, but it does appear that spatial visualization differences are most pronounced when some information must be preserved while the same or other information is being processed.