Abstract We present a thematic review and analysis of the variables affecting language learning motivation (LLM) (2008–2022). The second-language motivational self-system (L2MSS) model was found to be the most applied construct in measuring LLM. Complex systems theory was also another method gaining prominence in LLM research to explain the interactions between micro- and macro-structures surrounding the learner in influencing motivation. Other factors such as socioeconomic status, dialogism and anagnorisis were also identified as variables relating to LLM. For instance, research on dialogism and dialogue has indicated the role of conversation in shaping identity, motivation, and meaning for learners. However, our review found that much of the focus in LLM research has been on the L2MSS learning or teaching experience, while daily living has been largely neglected. We further conducted a reliability generalization meta-analysis. Our analysis found an average reliability of 0.84 (CI = 0.816–0.856), with 34% of reliability coefficients falling below the lower bound of CI. A meta-regression analysis revealed that 16% of the variance in the reliability coefficients was predicted by the number of items in the instruments. Questionnaires with an internal consistency below the lower bound of 0.816 had an average of 4.14 items, while the rest had an average of 5.71 items. We further found significant publication bias. Based on our findings, we suggest areas for future research in LLM.