INTRODUCTION In response to increasing demand for hospital beds, institution-wide clinical process redesign has been advocated for improving efficiency. METHODS This retrospective, before-after study involved five tertiary hospitals in Queensland, Australia and assessed effects of externally led redesign over 6 months within two hospitals, comprising ward-based innovations led by consultancy-led standardised processes, and internally led redesign over 25 months in one hospital which implemented medical assessment and planning unit, 23 h elective surgical ward and new bed management processes. The primary outcome measures were control chart changes in emergency department (ED) access block and overdue category 1 elective surgery waits over 3.5 years involving intervention hospitals and two control hospitals. RESULTS At one externally led redesign hospital, control charts indicated a decrease in ED access block outside control limits which coincided with the intervention, but this was not subsequently sustained. There were no special-cause variations seen in the other hospital. In contrast, at the internally led redesign hospital, there were two decreases in access block outside control limits during the intervention period, resulting in a decrease from a baseline average of 55% to a postintervention average of 22%. All hospitals showed declines in elective surgery waits with oscillations in data indicating the existence of special-cause factors other than redesign. CONCLUSION Internally led compared with externally led redesign led to superior and sustained improvements in ED access block as a result of major structural reforms that were driven by committed clinicians and managers and cut across departmental boundaries.