Factor analyses of two forms of the Graduate Record Examinations (GRE) Aptitude Test were undertaken to better understand the abilities affecting test performance. Results suggest that three global abilities--two verbal and one quantitative--are consistently tapped by the GRE Aptitude Test. Other less prominent dimensions--some of which appear to be specific to test forms--revealed aspects of the test related to item type, speededness, and the content of reading passages. Factor extension analysis was used to estimate the loadings on these operational test factors of new items from eight experimental tests administered with the operational forms. Statistical removal of the extended factors from the matrices of tetrachoric interitem correlations of the experimental tests and examination of residual relationships and amount of variance explained suggested that the experimental tests are, in general, adequately explained by the factors in the operational tests. There are, however, several dimensions in the experimental tests distinct from the factors underlying the operational forms. Recommendations based upon these findings are made concerning subtest length, item arrangement, and passage content. The factor analysis method is relevant for supplementing more traditional item classification and analysis techniques, and for the planned restructuring of the GRE Aptitude Test. (Author/ROF)