To successfully support healthcare decision-making, cost-effectiveness analysis (CEA) must consider and synthesise all (relevant) available evidence relating to the clinical effectiveness, health-related quality of life (HRQoL) and costs of the health technologies under scrutiny. Evidence based medicine tells us that statistical evidence synthesis of multiple individual patient level data (IPD) sources (e.g. IPD meta-analysis and its extensions) is the gold standard for deriving relative treatment effect estimates, one of the key parameters in any cost-effectiveness model.  Unfortunately the evidence base available to the cost-effectiveness modeller is often multifaceted and fragmented, comprising a mix of aggregate (AD, or summary level) and individual patient level data. This scenario poses a series of methodological issues and it is not uncommon for the analyst to end up collapsing the IPD into AD, with consequent loss of information, for use in a standard evidence synthesis model (e.g. meta analysis or mixed treatment comparison of AD).  Such a shortcut though impairs the analysis in many different ways, limiting our full understanding of the phenomenon being modelled and ultimately our ability to accurately assess ‘value for money’ beyond the simple 'average'. This paper explores the value of access to individual patient data for cost-effectiveness modelling, structuring the discussion of the topic around three interrelated questions.  First, what benefits can access to IPD bring to cost-effectiveness modelling?  Second, what are the challenges for the simultaneous statistically synthesis of AD plus IPD to derive input parameters for a cost-effectiveness model?  Third, what is the value of access to IPD compared to AD for cost-effectiveness modelling?  Using two different case studies, the above questions will be addressed and discussed in the context of the debate around CEA of individualised treatment decisions.