This study aims to investigate Taiwanese EFL students’ automaticity of reading in sentence processing, which includes identifying syntax and semantic features. Any communication task involves assessment and coordination of information from a multitude of perceptual, cognitive, and social domains. A foreign language learner must follow great amounts of rules and communicate the intended message unambiguously. However, humans are limited-capacity processors. In order to learn a language, learners need to integrate a number of different skills, and each of them must be practiced and routinized. In this study, twenty-three students participated in the experiment. They were divided into ‘High Proficient Learners’ and ‘Low Proficient Learners’. The results showed that ‘High Proficient Learners’ achieved a degree of automaticity with respect to processing form that the ‘Low Proficient Learners’ had not yet attained. Also, the greatest difficulty for students to automatize semantic processing was their insufficient vocabulary knowledge.