In this essay the researchers explore methods for synthesising multiple sources of evidence in health services research. The evaluation of policies or structural interventions in the healthcare system are complicated by the difficulty of linking the intervention to patient level outcomes. Proxy variables are often used in the place of important clinical outcomes such as mortality and quality of life. The researcher must therefore synthesise various forms of evidence from across a causal pathway that links the intervention to the outcomes of interest. The discussion is about whether the complex nature of a healthcare system hinders understanding of the relevant causal pathways, and whether such systems are amenable to modelling and the synthesis of multiple forms of evidence. An important distinction is drawn between the underlying phenomena that generalise across contexts and the complex data from which they are inferred. The CLAHRC team then explores the available methods a researcher might adopt and consider three steps: (i) the design of a causal model; (ii) the identification of the available evidence; and, (iii) the synthesis of multiple forms of evidence including quantitative and qualitative evidence. While many of the methods are well documented, further work is required to develop their use in combination, and understanding how to deal with different study designs, each with their own inherent biases, is still at an early stage. Nevertheless, there exist powerful tools for the synthesis of multiple forms of evidence in health services research.