Thirty-seven patients (41 shoulders) younger than 50 years of age (mean age, 43.7 years) underwent all-arthroscopic repairs of rotator cuff tears between 1994 and 2002. With a minimum 3-year follow-up (average 69 months, range 37-161 months), postoperative function was assessed by means of a UCLA scoring system, range of motion measurements, strength exam, and VAS pain score. Tears were categorized into 4 groups by type and size: partial undersurface tears, complete tears that were small (<1 cm), medium (1-3 cm), or large/massive (4-5 cm). All but 1 patient (97%) was subjectively satisfied and felt to be improved postoperatively. Average postoperative UCLA score improved significantly, and 95% of the shoulders achieved good or excellent postoperative results. Postoperative pain scores and strength evaluation also improved significantly. There was no significant loss of motion postoperatively and outcomes were independent of tear size and type. To date, there have been no revisions. These treatment results suggest that arthroscopic management in patients younger than 50 can achieve predictably good to excellent results regardless of tear size.