The use of machine translation has become a topic of debate in language learning, which highlights the need to thoroughly examine the appropriateness and role of machine translation in educational settings. Under the theoretical framework of task-technology fit, this explanatory case study set out to investigate the predictive role of machine translation fit, based on questionnaire responses obtained from a sample of 500 Chinese university EFL learners. Structural Equation Modeling approach was used to address the relationship between machine translation fit and learning performance, as well as the impact of technology and task characteristics on machine translation fit. The mediating role of machine translation literacy was further investigated in these relationships. The results showed that machine translation fit had a significant positive effect on learning performance. Both technology and task characteristics positively impacted machine translation fit. Meanwhile, machine translation literacy mediated the relationship between task characteristics and machine translation fit, but not technology characteristics and machine translation fit. This study has highlighted the significance of machine translation fit for language learning, providing suggestions and implications for integrating machine translation into language instructional practices.