During the Spring 2011 semester, a pilot study was conducted in a junior-level constraint-based modeling course at North Carolina State University where twenty-nine students were asked to model as many of the seven parts given in an assembly drawing of a device within a 110 minute class period. The main purpose of this pilot study was to determine the procedures necessary for this type of assessment in a classroom setting. The parts in the assembly ranged in complexity from a ball to a valve body. Students were given a ruler to measure parts on the B-size drawing and determine sizes of features based on the given scale (2:1). There was a positive relationship between the scores on the activity and the pace at which each student completed the parts. Only eight students modeled all seven parts in the assembly. Some of the students in the pilot study completely misinterpreted the 3D geometry of the parts. The researchers wondered if this was the result of insufficient practice reading drawings and/or the result of low spatial ability. During the Fall 2011 semester, the same assessment was administered to students at Riga Technical University and North Carolina State University who were enrolled in constraint-based/parametric modeling courses. Similar to the pilot study, students were asked to model the seven parts in an assembly drawing of a device within a 110 minute class period. For the current study, students also completed the Purdue Spatial Visualization Test – Visualization of Rotations (PSVT:R) to determine if the engineering graphics literacy assessment scores were related to spatial visualization ability. The researchers did not find significant differences between the universities on scores on the modeling test or scores on the PSVT:R. Overall, there was a significant correlation between scores on the modeling test and scores on the PSVT:R. Although other factors such as symbol recognition and understanding standards and conventional practices influence how well students read engineering drawings, it appears that spatial visualization ability plays a significant role it how well they visualize part geometry. One of the main concerns for conducting future studies is the ability to scale-up to handle more students. Although the rubric used in the pilot study and in this study delivered accurate assessments of the students’ modeling abilities, the time required to assess student work was very high. This potentially could prevent other faculty from using the instrument. The researchers plan on investigating alternative methods for accurately assessing student models such as automated programs for gathering the desired data from the models.