Concept maps measured a student's understanding of the complexity of concepts, and interrelationships. Novak and Gowin (1984) claimed that the continuous use of concept maps increased the complexity and interconnectedness of students' understanding of relationships between concepts in a particular science domain. This study has two purposes; the first one was to test this claim and examine how the repeated use of concept maps affected the complexity and interconnectedness of concepts independent of science subjects in elementary school, the second one was to compare the sensitivity of the Ruiz-Primo et al. (1997), and the Novak and Gowin (1984) grading systems for concept maps. The sample group consisted of 23 students including 14 male and 9 female students. We employed paired sample t-tests to answer the research questions, and found that the scores obtained for the fifth science unit was significantly different from the first one. Also, Novak and Gowin's (1984) scoring system was better than Ruiz-Primo et al. (1997) to evaluate complexity in students' thinking except for one of the units. We conclude that concept maps have the potential to measure change in complexity and interconnectivity of concept maps. Furthermore, repeated use of concept maps has the potential to increase the complexity and interconnectedness of student concept maps, and therefore improve their understanding of science independent of science content.