This study examined the relative effectiveness of a mnemonic method called the keyword method and a rote learning condition with 49 native speakers of Chinese. The students were from two classroom groups of intermediate-level students of equal age, proficiency, and learning background who were studying English at the SEAMEO Regional Language Centre in Singapore. We studied the students as intact classroom groups. We taught the keyword method to one group, and a rote learning method to the second group. After training, we tested the students' vocabulary learning twice: immediately following study of the test words (one day after the final practice session), and again seven days later. We found no statistically significant difference between groups on the immediate test scores (p > .05), but the keyword method group significantly outperformed the rote learning group on the delayed test (p < .05). The results accorded with most of the extensive research into the keyword method and suggested that the method can be used effectively in classroom environments with native speakers of Chinese who are experienced learners of English. However, the difficulty encountered in finding phonetic links between the languages limits the potential of the method for Chinese students learning English. A possible variation on the keyword method within the target language is described.