Abstract: Based on a small-scale Chinese-English comparable corpus, this paper uses a unified set of criteria in the identification and classification of the referring expressions and in the measurement of the accessibility of their referents in Chinese and English discourses.Results show that the overall pattern of distribution of both Chinese and English referring expressions in discourse corroborates with the prediction of the Accessibility Theory proposed by Ariel.The study also shows that, compared with the referential distance measured by discourse positions in which the anaphor and its antecedent appear, as employed in Ariel's analysis, the distance measured by the average number of other NPs intervening between the anaphor and its antecedent, as adopted in our analysis, not only proves to be a better indicator of the degrees of accessibility, but also can highlight the difference between Chinese and English,i.e., the average referential distance of Chinese referring expressions is shorter than their English counterparts.