Curriculum-Based Measurement (CBM) is emerging at middle/secondary school settings. The purpose of this study was to investigate the reliability of a concept maze task to assist middle school teachers in making accurate decisions regarding student content learning. A total sample of 240 middle-school students were randomly assigned to take a traditional maze or two different variations of the concept maze (examples or attributes). Results indicated that the concept maze using attributes as options was more difficult and more stable relative to the other two types of mazes. Alternate form and test-retest analyses of the concept mazes resulted in moderate correlations. Results are discussed in the context of extending CBM into middle and secondary classrooms as a tool that teachers can use to monitor the learning progress of their students.