As outlined by Benjamin Bloom, students working within a mastery learning framework must demonstrate mastery of the core prerequisite material before learning any subsequent material. Since many learning systems in use today adhere to these principles, an important component of such systems is the set of rules or algorithms that determine when a student has demonstrated mastery. A relevant issue when discussing mastery learning is its durability--in particular, we are interested in the relationship between different mastery learning thresholds and the forgetting of the learned material. As such, in this study we investigate this question using a large data set from the ALEKS adaptive learning system. Applying a quasi-experimental design, we find evidence that, while a higher mastery threshold is initially associated with a higher rate of knowledge retention, after several weeks this difference has largely disappeared. [For the full proceedings, see ED623995.]