Objective. To examine the effect of strenuous exercise on the magnetic resonance imaging (MRI) characteristics of the rotator cuff tendon. A second objective was to define an optimal time to image the rotator cuff and possibly eliminate exercise-induced false positives. Design and patients. Five male subjects from 24 to 38 years old with normal rotator cuffs by history, physical examination, and screening MRI underwent a rotator cuff exercise session on the Biodex System 2 (Biodex, Shirley, New York). The exercise sessions were followed by sequential MRI scans of the exercised shoulder. These were performed immediately and at 8 h and 24 h after exercise. Results and conclusions. The rotator cuff tendon and subacromial-subdeltoid bursal signal remained unchanged from the pre-exercise through the 24-h post-exercise scans. The rotator cuff muscle signal was increased in five of five subjects on the immediate post-exercise fat-suppressed T2-weighted images. This signal returned to baseline by the 8-h scan. Positive findings of rotator cuff pathology on MRI after strenuous athletic activity should not be discounted as normal exercise-induced changes. Also, diagnostic MRI scanning may take place after a practice session without an increased risk of false positives.