BACKGROUND: The natural history of small, symptomatic rotator cuff tears is currently unclear. The purpose of the present study was to assess the clinical and structural outcomes for a consecutive series of patients with symptomatic, isolated full-thickness supraspinatus tears who had been offered rotator cuff repair but declined operative treatment. METHODS: In the study period, twenty-four patients with isolated full-thickness supraspinatus tears that had been diagnosed by means of magnetic resonance arthrography were offered rotator cuff repair and elected nonoperative treatment. The twenty men and four women had an average age of fifty-two years at the time of diagnosis. At a median of fortytwo months after the diagnosis, all patients were reexamined clinically according to the Constant and Murley scoring system and all shoulders underwent standard magnetic resonance imaging. RESULTS: At the time of follow-up, the mean subjective shoulder score was 74% of that for a normal shoulder and the mean Constant score was 75 points (relative Constant score, 86%). The mean rotator cuff tear size did not change significantly over time (95% confidence interval, 0.51 to 1.12). In two shoulders, the tear was no longer detectable on magnetic resonance imaging, in nine shoulders the tear was smaller than it had been at the time of the initial diagnosis, in nine patients the tear had not changed, and in six patients the tear had increased in size. There was a slight but significant progression of fatty muscle infiltration of the supraspinatus, but no patient had fatty infiltration beyond stage 2 at the time of the latest follow-up (95% confidence interval, 0% to 14%). CONCLUSIONS: In a consecutive series of patients who had been offered repair of an isolated, symptomatic supraspinatus tear, the refusal of operative treatment resulted in surprisingly high clinical patient satisfaction and no increase of the average size of the rotator cuff tear 3.5 years after the recommendation of operative repair. This study confirms that the size of small rotator cuff tears does not invariably increase over a limited period of time. Distinguishing tears that will increase in size from those that will not needs further study. Background: The natural history of small, symptomatic rotator cuff tears is currently unclear. The purpose of the present