A Cognitive Grammar (CG) analysis of linguistic constructions has been claimed to be beneficial to second language teaching. However, little empirical research has been done to support this claim. In this study, two intact classes of Chinese senior high school students were given a 45-minute review lesson on the usages of the English simple present tense. Instruction for the experimental class was based on Langacker's cognitive grammar analysis that highlighted the common motivation linking various usages, while that for the control class followed traditional teaching method. Results showed that the learners in the CG approach did not perform significantly better than the control group both on a grammaticality judgment and error correction task, and on a fill-in-the-blank task, though the CG approach did improve students' performance considerably. The author argued that four reasons might have diminished the benefits of the CG approach: 1) the selection of the samples was inappropriate; 2) the CG approach ran counter to learners' expectations about grammar; 3) the presentation of the CG treatment was too abstract, and linguistic terminologies new to the students were introduced, which made the lesson hard to follow; 4) instructional time was too limited. Therefore more classroom research is needed to substantiate the claimed pedagogical benefits of the CG approach.