In many mixed-criticality systems, tasks of differing criticality levels execute upon a common computing platform. Within such systems, it is imperative that communication between tasks, and between tasks and shared computing resources, do not impede upon the correct functionality of highly critical tasks. To evaluate the performance of communication mechanisms that offer this functionality, a number of attributes of the mechanism may be assessed. Currently, there exists no agreed upon method of evaluating such mechanisms, and so comparing the performance of differently developed mechanisms is difficult. Within this paper, we identify attributes that characterise the performance of mixed-criticality communication, offer suggestion as to how they may be comparably evaluated, and outline an evaluation framework that would simplify the process of comparably evaluating a communication mechanism with these scenarios.