This paper reports on the creation of an instrument for use by instructors, students, or researchers to identify, measure or promote critical thinking (CT) in online asynchronous discussions (OADs). Four models of CT were reviewed, synthesised, analysed, and evaluated to clarify the construct. Indicators of specific cognitive processes related to CT were identified, and subsequently retained, rejected or modified to show how the construct might be operationalised in real contexts of use. Subsequent empirical testing of the instrument for the analysis of a transcript of an OAD revealed that while the instrument was valuable in identifying and measuring CT in the OAD, issues of practicality need to be addressed. Implications for research and practice are presented. (Contains 3 tables.)