Background: Hygiene promotion is a cornerstone of humanitarian response during infectious disease outbreaks. Despite this, we know little about how humanitarian organisations design, deliver or monitor hygiene programmes, or about what works to change hygiene behaviours in outbreak settings. This study describes humanitarian perspectives on changing behaviours in crises, through a case study of hygiene promotion during the 2014-2016 Liberian Ebola outbreak. Our aim was to aid better understanding of decision making in high-stress situations where there is little precedent or evidence, and to prompt reflection within the sector around how to improve and support this. Methods: We conducted in-depth, semi-structured interviews with fourteen purposively-sampled individuals (key informants) from international organisations involved in hygiene behaviour change during the outbreak. Through thematic analysis we identified the decisions that were made and processes that were followed to design, deliver and monitor interventions. We compared our findings with theory-driven processes used to design behaviour change interventions in non-outbreak situations. Results: Humanitarians predominantly focussed on providing hygiene products (e.g. buckets, soap, gloves) and delivering messages through posters, radio and community meetings. They faced challenges in defining which hygiene behaviours to promote. Assessments focused on understanding infrastructural needs, but omitted systematic assessments of hygiene behaviours or their determinants. Humanitarians assumed that fear and disease awareness would be the most powerful motivators for behaviour change. They thought that behaviour change techniques used in non-emergency settings were too 'experimental', and were beyond the skillset of most humanitarian actors. Monitoring focussed on inputs and outputs rather than behavioural impact. Conclusions: The experiences of humanitarians allowed us to identify areas that could be strengthened when designing hygiene programmes in future outbreaks. Specifically, we identified a need for rapid research methods to explore behavioural determinants; increased skills training for frontline staff, and increased operational research to explore behaviour change strategies that are suited to outbreak situations.