Six different 20-minute spatial tests were given to a group of 254 male first-year college students. Correlations were computed between the several tests and between each test and grades in mathematics and engineering drawing. Intercorrelations between the tests ranged from .34 to .58 with a median of .48. Validity coefficients ranged from .16 to .40 with mathematics and from .37 to .68 with drawing grades. Factor analysis indicated a grade factor and two closely related spatial factors (correlation about .7). These two factors were best defined by tests specially designed to increase their factorial "purity." These "pure" tests had the lowest validities of the six tests. One of the spatial factors was found to be related more closely than the other factor to each of the criterion scores. The appendix contains the table of interrcorrelations between the six tests and two criterion scores. It lists the centroid factor loadings and the loadings after an orthogonal rotation to a positive manifold. It also includes the directions pages for each of the six tests.