If it is to be argued that self-access language learning centres (SACs) provide an efficient and effective alternative or complement to more traditionally accepted modes of language learning and teaching, it remains a serious concern that there is no research-based framework specifically developed for their evaluation. It is within the context of a study into the development of such a framework that this paper considers the issue of evaluating learning gain in a SAC. Experience and research suggest that the issue of learning gain is crucial to any discussion of SAC evaluation. There is, however, a tension between the perceived need to include learning gain as a focus in an evaluation of a SAC's effectiveness, and a recognition of there being practical problems in so doing. After briefly examining relevant literature, this paper will discuss data derived from the aforementioned study. It will present a number of reasons for attempting to evaluate learning gain, before going on to identify potential difficulties involved and discussing possible ways of achieving this. It will finally suggest that, in order to evaluate learning gain within a self-access language learning context, there is a need to re-conceptualize what is perceived as learning gain and to look to the learner to identify evidence of such. (Contains 1 table and 1 figure.)