This study examined the potential use of keystroke dynamics to create keyprints (typing fingerprints) to authenticate individuals in online courses. Previous studies have determined that keystroke typing patterns are unique; in addition to replicating these findings, this study explored best practices for implementing keyprint signatures in situations other than simple password or key phrase verification. While authentication can be difficult when attempting to correctly identify individual users, the results of this study indicate that keyprints can reliably indicate negative cases (a typing sample that is likely not the intended student). The results of this study suggest that complete keyprint signatures are better than keyprint profiles (reduced keyprints based on unique typing patterns only). Results also suggest that keyprint signatures are most reliable when a variety of methods are used to obtain keyprint data (i.e., both copy typing and free typing). Small comparison samples increase the challenge of identification but do not prevent the system from identifying negative cases.