Using Kuhn and McPartland's approach, 116 Chinese college students were recruited and asked to write as many sentences as possible beginning with "Chinese...," "Americans...," and "Japanese...." The population of sentences consisted of 258 adjectives, of which 96 described Chinese, 53 described Americans, and 109 described Japanese. Next, the first ten adjectives with the highest frequencies describing Chinese, Americans, and Japanese, respectively were selected for the second step of data collection in that students were asked if they would accept these adjectives describing Chinese, Americans, and Japanese. If they did, they were further asked to provide examples or illustrations to support their agreement. Based on the top 30 adjectives and the qualitative data collected at the second step, students' perceptions of Chinese, American, and Japanese were analyzed. The results of the data analysis discovered that Chinese students are, to some extent, stereotyped toward American and Japanese, and their understanding of Chinese people is not accurate, either. The results confirm that media to certain extent are responsible for the shaping of some of the students' stereotypes. Evidently, the results of this study suggest that the Chinese students' stereotypes will affect their intercultural communication with Americans and Japanese. Although this study does not provide tactics on the reduction or elimination of stereotypes in the process of intercultural communication, it provides meaningful clues for future research in how stereotypes may be reduced by way of intercultural contact or training. This research has two major limitations. First, all the subjects are Chinese. Had American and Japanese samples been selected for a comparative investigation, more meaningful and insightful results could have been obtained. Second, because of the space constraints, verbs and nouns were not analyzed. (Contains 3 tables and 1 footnote.)