The challenges associated with the responses to crises1 that affect society have been studied for almost a hundred years (Lindell, 2011; Scanlon, 1988). Researchers from various disciplines continue to remind us that response management is characterized by complexity, and that response organizations must be prepared for the unexpected and be able to adapt to new conditions (see e.g. Cutter et al., 2010; Mendonça and Wallace, 2004; Wachtendorf, 2004; Weick and Sutcliffe, 2015). In this context, it may be necessary to rely on what practitioners and academics alike describe as improvisation. Improvisation can be seen as a tool to deal with situations that require action without planning (see e.g. Moorman and Miner, 1998b), or where the ability to develop a plan is hindered by uncertainty (see e.g. Crossan et al., 2005). Most importantly, improvisation is a well-known response to crises (see e.g. Kendra and Wachtendorf, 2006; McEntire et al., 2013; Mendonça and Wallace, 2004), leading Tierney (2002, cited in Kendra and Wachtendorf, 2006, p. 1) to conclude that “if an event doesn’t require improvisation, it is probably not a disaster”. This paper is a critical analysis of the concept of improvisation as it is used in the scientific literature, and how it is related to capability in the context of crisis management. Although much of the crisis management research has focused on first responders, such as police and fire brigades, recent events such as hurricane Katrina in 2005, the eruption of Eyjafjallagökull in 2010, and the Great East Japan earthquake in 2011 that led to the Fukushima Daiichi nuclear disaster, have illustrated the transboundary nature of modern crises (Ansell et al., 2010). When the consequences of a crisis are spread across geographical, administrative and sectorial boundaries (e.g. from power distribution to transport, to health care) it becomes much harder to manage. This implies that in major crises, many organizations that are not traditionally seen as first responders become very important for managing its consequences. In this article, we pay specific attention to such contexts in our analysis of the meaning and role of improvisation. One key reason why modern crises may be transboundary, for example, crossing both geographical and functional borders, is the increased interconnectedness of critical infrastructures (CIs). A disturbance in one infrastructure can easily spread to the next, giving rise to a cascade of failures (Rinaldi et al., 2001; de Bruijne and van Eeten, 2007). The extension of the consequences beyond the system in which the disturbance originated can intensify the strain on society. Additionally, cascading effects can make it much harder to manage the initial event, and to restore vital societal functions.