The War in Syria displaced over 5.5 million Syrians to neighbouring states and created unprecedented humanitarian need. However, only some international organizations (IOs) and international non-governmental organizations (INGOs) made disruptive changes to their service delivery and rulemaking and addressed the unique, emerging health needs of Syrian refugees and migrants. The predominant expectations of scholars are that under severe time and security constraints, aid workers will have difficulty learning and changing their behaviors. We anticipate they will respond to new problems in a new context with the activities and strategies they used elsewhere or make only small adaptive changes to what they do. This paper examines moments when a few humanitarian organizations working amongst Syrian refugees in Lebanon and Jordan bucked this trend and asks if existing theories of institutional change and organizational behavior can capture these moments. Based in 10-months of political ethnography and over 120 interviews with aid workers, I argue that organizations have more control over their response to external pressures than past studies indicate and existing theories capture, and that this obscures deliberate, fast-moving, and disruptive efforts by non-state actors to govern and exert power and influence over migrants in the Middle East. Building on theories of sociological institutionalism and responding to what Streeck and Thelen describe as a “conservative bias” toward gradualism in the literature, I show that actors inside of international organizations and below centralized headquarters can generate sufficient contestation surrounding humanitarian ‘need’ to trigger ruptures in the logics of global humanitarian institutions. I find that variation in aid worker capacities to trigger change is a function of organizational design.