This article provides a comprehensive synthesis of research on language attrition to date, with a view to establishing a theoretically sound basis for future research in the domain of second language (L2) attrition. We identify the variables that must be tracked in populations who experience language loss, and we develop a general model for the assessment of the processes involved. This critical review suggests that future research in this domain should establish baselines for attainment against which to measure attrition, and that learners must be compared to themselves in longitudinal designs that involve periodic assessment of both linguistic and extralinguistic factors. In the proposed model, populations are defined as sets of variables, which are subject to change following shifts in discrete time periods in the general process of acquisition and attrition. A working model is elaborated for the assessment of L2 attrition and retention, which, we hope, might encourage additional work in this area. (Contains 4 notes, 1 table, and 2 figures.)