To inform policy decisions economic evaluation (EE) studies require the systematic identification and (quantitative) synthesis of the relevant evidence base on the clinical effectiveness, quality of life (QoL) and costs associated with the use of competing health technologies.  Existing methods for EE are linked to principles of evidence-based medicine and, as such, are geared primarily towards the evaluation of pharmaceuticals. Some authors have claimed that medical devices (MDs) cannot be evaluated using the same principles. We take the opposite viewpoint and argue that used within the right evaluative framework existing EE methods are indeed appropriate to assess the cost effectiveness of therapeutic MDs.  What makes the (economic) evaluation of MDs challenging is the fact that the quantity, quality, and characteristics of the evidence base around them, is often fragmented, heterogeneous and associated with high levels of uncertainty. In these circumstances it is important to acknowledge the value of eliciting and quantitatively summarising physicians and other experts’ beliefs regarding the effectiveness and resource use demands associated with MDs already in use.  Using real life examples this paper shows how a Bayesian stepwise iterative approach has helped address some of the challenges associated with the EE of MDs, while guiding policy decisions regarding technology adoption, research funding and design. Relevant steps include: (a) identification of existing evidence base and elicitation of experts’ beliefs on clinical effectiveness , QoL and costs - i.e. “a priori evidence base”; (b) quantitative synthesis of this a priori evidence base to inform the parameters of an EE model; (c) initial estimation of the model; (d) assessment of the economic value of conducting further research (VoI); (e) collection of new patient level data (PLD) in a pilot study; (f) new evaluation of the EE model updating the prior estimates using primary PLD; (g) further VoI analysis.