Purpose The purpose of this study was to compare the clinical outcome of arthroscopic rotator cuff repair with metal and biodegradable suture anchors. Methods Arthroscopic rotator cuff repair was performed in 110 patients with a full-thickness rotator cuff tear. They were divided into 2 groups of 55 patients each, according to suture anchors used: metal anchors in group 1 and biodegradable anchors in group 2. Results were evaluated by use of the Disabilities of the Arm, Shoulder and Hand (DASH) and Work-DASH self-administered questionnaires, as well as the Constant score normalized for age and sex. On analyzing the results at 2 years' follow-up, we considered the following independent variables: baseline scores; age; sex; arm dominance; location, shape, and retraction of cuff tear; fatty degeneration; treatment of biceps tendon; rotator cuff repair technique (anchors or anchors and side to side); and number of anchors. Univariate and multivariate statistical analyses were performed to determine which variables were independently associated with the outcome. Significance was set at P < .05. Results Of the patients, 9 (8.2%) were lost to follow-up. Comparison between groups did not show significant differences for each variable considered. Overall, according to the results, the mean DASH scores were 17.6 ± 17.2 points in group 1 and 22.8 ± 19.9 points in group 2; the mean Work-DASH scores were 24.9 ± 28.1 points and 22.5 ± 24.1 points, respectively; and the mean Constant scores were 104 ± 20.5 points and 98.6 ± 14.3 points, respectively. Differences between groups 1 and 2 were not significant. Univariate and multivariate analysis showed that only baseline score, age, tear location, and fatty degeneration significantly and independently influenced the outcome. Conclusions At a short-term follow-up, differences between arthroscopic repair of full-thickness rotator cuff tears with metal and biodegradable suture anchors were not significant. Level of Evidence Level I, high-quality randomized controlled trial with no statistically significant differences but narrow confidence intervals.