Updating, improving and spreading the evidence base for healthcare practices has proven to be a challenge of considerable magnitude - a wicked, multi-dimensional problem. There are many interlinked factors which determine how, why and whether any particular implementation effort or intervention succeeds. Soft Systems Methodology (SSM), strongly grounded in systems ideas and complexity science, offers a structured, yet flexible process for dealing with situations that are perceived as problematical and in need of improvement. The aim of this paper is to propose the use of SSM for managing change in healthcare by way of addressing some of the complexities. The aim is further to illustrate examples of how SSM has been used in healthcare and discuss the features of the methodology that we believe can be harnessed to improve healthcare.SSM is particularly suited for tackling real world problems that are difficult to define and where stakeholders may have divergent views on the situation and the objectives of change. SSM engages stakeholders in a learning cycle including: finding out about the problematical situation, i.e. the context in which the problem exists, by developing a rich picture of the situation; defining it by developing conceptual models and comparing these with the real world; taking action to improve it by deciding on desirable and feasible improvements; and implementing these in an iterative manner. Although SSM has been widely used in other sectors, it has not been extensively used in healthcare. We make the case for applying SSM to implementation and improvement endeavours in healthcare using the example of getting clinicians at the hospital level to use evidence-based guidelines.Applying SSM means taking account of the multi-dimensional nature of care settings, and dealing with entrenched and unique contexts, cultures and socio-political ecosystems - precisely those that manifest in healthcare. There are gains to be made in appreciating complexity and facilitating contextualization of interventions, and by approaching improvements in an iterative learning cycle.