While problem solving as an instructional technique is widely advocated, educators are often challenged in effectively assessing student skill in this area. Students failing to solve a problem might fail in any of several aspects of the effort. The purpose of this research was to validate a scaffolded technique for assessing problem solving in science and social studies at the middle school level. This technique attempts to isolate three aspects of problem solving (data collection, analysis and display, and interpretation) and to measure each aspect separately. Problem solving measures were developed in both science and social studies. These were administered both fall and spring to determine student skill in problem solving and to measure growth in problem solving skill over time and differential skill across grades (6 through 8). Segmented tasks were scaffolded between segments to circumvent the interdependency of elements of the problem solving process. It was determined the measures were successful in supporting students who had difficulty across segments within a single problem solving task and student problem solving skills could be evaluated effectively using the results of the measure.