Background Physiotherapists increasingly manage shoulder referrals in place of orthopaedic doctors. Better understanding the agreement between these professionals will help inform the safety, quality and potential costs of these care models. Objective To establish the level of agreement between a physiotherapist and an orthopaedic surgeon regarding diagnosis, management and corticosteroid injection, in a representative sample of orthopaedic shoulder referrals. Design Blinded inter-rater agreement study. Method 274 public orthopaedic shoulder patients were independently assessed by a physiotherapist and an orthopaedic surgeon. Management, subacromial corticosteroid injection, diagnosis and investigation decisions were compared using inter-rater reliability statistics. Results Agreement between the physiotherapist and the orthopaedic surgeon was near perfect for surgical versus nonsurgical management (Gwets agreement coefficient AC1 = 0.93, 95%CI: 0.90-0.93), safety of injection (AC1 = 0.85, CI: 0.79-0.91) and investigations requested (AC1 = 0.87, CI: 0.83-0.91); substantial for the presence of subacromial pain (AC1 = 0.74, CI: 0.66-0.81) and diagnosis (AC1 = 0.72, CI: 0.66-0.78); and moderate regarding delivery of subacromial corticosteroid injection as an immediate treatment (AC1 = 0.48, CI 0.33-0.53), with the physiotherapist less inclined to select corticosteroid injection as the first intervention. Conclusion In this study a physiotherapist with prescribing and injection training made decisions analogous to those of an orthopaedic surgeon at initial consultation for orthopaedic shoulder pain, including the safe identification of patients for subacromial injection, without prior screening of referrals by orthopaedic doctors. Trial registration Australia and New Zealand Clinical Trials Registry, number 12612000532808.