Spatial visualization skills are essential in many engineering and technology fields. These skills are especially important in hand sketching and computer-aided design (CAD) of engineering graphics, when it is necessary to visualize and represent three-dimensional (3D) objects and assemblies. Educators and researchers have developed various types of tests to assess students' abilities in spatial visualization tasks. Conventional visualization tests usually use axonometric drawings, mostly isometric projections, to represent 3D objects. However, isometric drawings are dimensionally distorted and lack many features present in a realistic view of 3D objects. They are also prone to drawing errors, which, when combined with the inherent distortions of isometric drawing, may lead to misjudgment of students' visualization abilities. In order to accurately assess students' visualization ability, realistic views are necessary in spatial visualization tests. To investigate the effects of 3D views, the author conducted a comparative study of a popular spatial visualization test given to the first-year graphics and CAD classes at a minority community college. The objects in the Purdue Spatial Visualization Test-Visualization by Rotations (PSVT-R) were recreated with 3D solid modeling CAD software to show more realistic views. The results from both the original test in isometric views and the same test with 3D solid model views are statistically analyzed. Also, a more detailed discussion is presented of the advantages of using solid modeling in spatial visualization tests, and the drawbacks of the conventional test using isometric drawings.