Drawing a paper-based concept map gives students more freedom to express their mental model than digital concept mapping tools. However, this freedom can hinder a uniform structure and make determining the structural quality of students’ mental models more difficult. In this dual-study paper, we develop and determine the reliability of our Width, Depth and Strength Tool (WiDeST) for assessing paper-based concept maps in secondary and tertiary education. In the first study, 157 secondary education students created 1377 concept maps in a longitudinal design over 24 weeks. The first study's results indicate that WiDeST is reliable, with an Omega Total of 0.81. Test-retest stability (ICCk2) ranges between 0.72 and 0.84. To test whether WiDeST remained reliable in tertiary education, we undertook a second study in which 80 students created 80 concept maps. The second study's results show that WiDeST is reliable with an Omega total of 0.70. WiDeST remained reliable while the structural complexity of the mental models increased from secondary education to tertiary education.