This paper investigates members’ practices at Drug and Therapeutic Committee (DTC) meetings. It describes situated practices involving textual and other forms of ‘evidence’ considered by committee members at two DTC sites dealing with the managed entry of a single pharmaceutical product (clopidogrel). It investigates the locally situated textual activities of committees and members; issues that evidence-based medicine (EBM) programs ignore in their assumption that decision-making can, and should, be a straightforward logical process involving the reading, understanding, and evaluation of scientific data unambiguously from a text. It describes the actual usage of written textual information by committees and individuals within them, as sophisticated reflexive workplace activities. Taking an ethnomethodological orientation, this paper shows how the same clinical study can result in different activities and discussions. This is not only because of different ‘readings’ by personnel in different places, at different times, and with different local concerns, but also due to the representational formats in which the study exists as a text document and object, i.e., that ‘evidence’ does not remain uniform across various sites.