This paper presents the results of a systematic review of the empirical literature, reported in research journals, on Mathematics Learner Identity (MLI). In the mathematics education research field, arguments have been made that the conceptualization of Mathematics Identity (MI) is inconsistent and this makes the literature as a whole incoherent. This study aims to summarize how the concept has been employed and to develop an integrative model for analyzing concept-definitions and concept-operationalizations of MLI. An analysis of 69 papers revealed 3 main dimensions that characterized how identity is conceptually defined in any research literature (social/subjective, enacted/representational and change/stability) and 5 main categories that describe how the literature has implemented these dimensions operationally (identity as individual attributes; identity as narratives; identity as a relationship with specific practices; identity as ways of acting; and identities as afforded and constrained by local practices). An emphasis on representational aspects of identities, particularly during higher education, and enacted and practice-related identities during primary and secondary education research was found. The discussion shows how the field would be clarified if studies made their choices of conceptualization clear in terms of these dimensions and categories, and if research as a whole becomes more aware of restrictions/limitations in identity research, particularly at different points of the educational trajectory of students.