We examined the outcomes of patients who underwent surgical treatment for rotator cuff tears to test the hypothesis that patients who did not smoke would have better postoperative scores than smokers. Medical charts were reviewed for patients who underwent rotator cuff repair between 1990 and 1993. We examined age, smoking status, workers' compensation status, and the size of the rotator cuff tear to determine the effect of scores on the UCLA questionnaire and a subjective pain assessment. There were 95 smokers and 129 nonsmokers. Mean preoperative UCLA scores for smokers and nonsmokers were 15.9 and 17.6, respectively; mean postoperative scores were 25.0 and 32.0, respectively. Nonsmokers had a significantly greater increase in total UCLA scores than smokers (P <.0001) and significantly higher improvement in pain scores, and more nonsmokers were classified as having good or excellent results based on the UCLA rating. On the basis of our data, nonsmokers undergoing rotator cuff repair have greater improvement of pain and better results postoperatively than smokers.