The Airborne Collision Avoidance System (ACAS) has been in use worldwide for many years as a 'last resort' means of preventing aircraft mid-air collision. Although its efficacy was predicted via extensive simulation studies, measurement of actual collision risk reduction in the airspace is impractical. Furthermore, ACAS had not been subjected to contemporary safety assessment practices. Upon its mandated deployment in European airspace, ACAS therefore presented a unique opportunity to apply state-of-the-art thinking on air traffic management safety cases (the so-called 'success and failure' approach) to a mature operational system. The paper describes the background to the safety case, the safety assessment process that underpinned it by synthesising hitherto missing evidence, and some of the safety issues it has revealed.