Tutoring gives tutors the opportunity to engage in interactive strategies that help them to assess a tutee’s understanding. However, tutors without teaching experience often do not engage in interactive strategies and, thus, have difficulty with accurately assessing a tutee’s understanding. We conducted an experiment with 39 tutor-tutee dyads to test whether tutors who received training in interactive strategies would become more interactive and more accurate in assessing a tutee’s understanding. Results showed that trained tutors provided a more interactive style of tutoring than untrained tutors. However, due to being more interactive, trained tutors produced less accurate assessments than untrained tutors. This suggests that changing the style of tutoring to implement interactive strategies puts a high burden on a tutor’s cognitive capacity. Hence, there is obviously little cognitive capacity left that could be used to assess a tutee’s understanding. Training methods that automate strategy use might enhance a tutor’s assessment accuracy.