To date, no studies have been published that have assessed the optimal position of sling immobilization after anatomic total shoulder arthroplasty for glenohumeral osteoarthritis. Thirty-six patients undergoing anatomic total shoulder arthroplasty for osteoarthritis were randomized to a neutral rotation sling versus an internal rotation sling. The primary outcomes assessed included the Disabilities of the Arm, Shoulder and Hand score; Western Ontario Osteoarthritis of the Shoulder score; Single Assessment Numeric Evaluation score; visual analog scale (VAS) scores for pain and satisfaction; compliance ratings; and radiographic and range-of-motion measurements. Primary outcomes were assessed at baseline and postoperatively at 2 weeks, 6 weeks, 3 months, 6 months, and 1 year. All patient-determined outcome scores for both groups revealed statistically significant improvements (P < .0001) from enrollment to final follow-up. There were statistically significant advantages to the neutral rotation sling group compared with the internal rotation sling group when we evaluated the improvements in (1) active external rotation (42° vs 25°, P = .03), (2) passive external rotation (44° vs 26°, P = .02), (3) passive horizontal adduction (7.7 cm vs 3.7 cm, P = .05), and (4) pain relief with passive adduction (VAS score, 6.2 cm vs 3.5 cm; P = .002). There was a trend toward greater improvements in the neutral rotation sling group when we measured (1) active horizontal adduction (8.3 cm vs 2.9 cm, P = .06) and (2) active internal rotation behind the back (18 cm vs 11.1 cm, P = .09). At 2 weeks, the neutral rotation sling group had significantly less night pain than the internal rotation sling group (mean VAS score, 18 mm vs 34 mm; P = .047). Neutral rotation sling use after anatomic total shoulder arthroplasty resulted in statistically significant improvements in external rotation and adduction, as well as decreased night pain, compared with an internal rotation sling.