The purpose of this multi-centre study was to assess external rotation in a large cohort following reverse shoulder arthroplasty (RSA) at a minimum follow-up of 2years, and identify factors that influence postoperative and/or net-improvement of external rotation. The authors retrospectively reviewed records of 743 RSAs performed between January 2015 and August 2017 by 16 surgeons that participated in a large national society symposium; 193 (25.7%) were lost to follow-up, 16 (2.1%) died, and 33 (4.4%) were revised with implant exchange, which left 501 available for assessment at 2.0-5.5years. Pre- and post-operative active forward elevation, active external rotation (ER1), active internal rotation (IR1) and constant score (CS) were collected. Regression analyses were performed to determine associations of patient demographics, surgical and implant parameters, rotator cuff muscles status and radiographic angles with ER1. Multivariable analyses revealed postoperative ER1 decreased with age (β, - 0.35), increased with lateralisation shoulder angle (LSA) (β, 0.26), and was better in shoulders operated by the antero-superior (AS) approach (β, 11.41), but worse in shoulders with absent/atrophic teres minor muscles (β, - 10.06). Net-improvement of ER1 increased with LSA (β, 0.39), was better with inlay stems (β, 8.33) and BIO RSA (β, 6.22), but worse in shoulders operated for primary OA with rotator cuff (RC) tears (β, - 16.26), for secondary OA due to RC tears (β, - 16.06), or for mRCT (β, - 18.96). This large multi-centre study revealed that, at a minimum of 2years following RSA, ER1 improved by 16.1°. Postoperative ER1 was better in shoulders which had normal or hypertrophic teres minor muscles, were operated by the AS approach, or with greater LSA. Net-improvement of ER1 was better in shoulders with inlay stems, with BIO RSA, or with greater LSA, but worse in shoulders with rotator cuff deficiency. IV.