The large choice of protocol configurations available in this task has been proven to be very useful in assessing neurobehavioral bases for learning and memory. In the procedure to assess spatial working memory, each arm of the maze is baited and repeat entries into an arm that has already been visited constitute a working memory error. In the procedure to assess reference memory, only some arms are baited at the beginning of the session. First entry into a non-baited arm constitutes a reference memory error, and repeated entries to baited and non-baited arms are defined as a working memory error.