To collect data on current management and outcome of acute severe acromioclavicular joint (ACJ) injuries to inform the knowledge base, design and conduct of future research and explore the patient and injury features predicting surgical management. A prospective cohort study was conducted by two trainee collaboratives of acute Grade III to VI ACJ injuries presenting to 12 hospital trusts. 54 Patients were recruited within four weeks of injury regardless of treatment type over a one-year period. Patient reported outcomes and healthcare resource use were collected at six and 12 months post injury. Accounting for Rockwood grade, age, gender and dominant arm injury, the operative group had a statistically lower Oxford Shoulder Score (OSS) at baseline (10.8 vs 25.3, <0.0001) and at six months (37 vs 43.8, p = 0.024) than the non-operative group. There was no statistical difference at 12 months (40 vs 44, p = 0.205). The odds of operative management were inversely correlated with baseline OSS (OR 0.89, p = 0.014). There was no clear preferential method of fixation in the ten patients that received surgery. Future research needs to take into consideration the transitory population of working age, on response to follow up. A future randomised clinical trial (RCT) should be multicentred, pragmatic and stratified for baseline function and injury grade at randomisation.