Mandarin Chinese is always classified as a topic-prominent language (Li and Thompson 1975). One of the characteristics of a topic-prominent language is that pronouns may drop since speakers and addressees know what they are talking about. It is this feature that makes topic-prominent languages or pronoun-drop languages interesting, for the dropped pronoun or the zero pronoun can be controlled by the topic in the previous discourse not just in the local sentence. Interestingly, there are three levels or layers in Chinese speech (Li, Ing Cherry 1985, Chu 1991), which causes foreigners to make some mistakes when they make Chinese sentences and paragraphs because they might not know how to use noun phrases, pronouns, and zero pronouns in a proper way. That’s why Chinese discourse grammar is important in teaching/learning Mandarin Chinese as a foreign language or as a second language. Jyun-Gwang Chen (2008) found the specific rules of the third person singular tā他 in Chinese discourse grammar from the linguistic database. His study is meaningful on the view of Chinese discourse grammar. Chen (2008) reported that the distributions of zero pronoun are the most unmarked and the most prevailing way of anaphora in Chinese. Pronouns and noun phrases are used markedly as event markers in Chinese speech (Chen 2008). However, it seems that Mandarin Chinese in Taiwan has changed a great deal since its establishment as national language in the early years of the republic and is still changing. Non-human pronoun它tā, for instance, is used more frequently nowadays. The purposes of the present study are to find out if there is language change of Mandarin Chinese in Taiwan, involving the use of pronoun它tā, and to what extent has the rules of its use been changed. In addition, we hope to discover what are the sociolinguistic factors involved in the changes. There are English-Chinese and Chinese-English translation exercises in our study. We elicit subjects to translate English pronoun it into Chinese in order to check if the zero anaphoric system of Mandarin Chinese has changed. Besides, an investigation of discourse database has been done to reconfirm our study purposes. The results of the translation exercises support our claim that Mandarin Chinese in Taiwan has changed. Currently, a few speakers keep using zero pronoun more often than non-human pronoun它tā, while other speakers tend to use less zero pronoun and more and more non-human pronoun它tā. A close examination of the discourse database also shows that the usage of non-human pronoun它tā in subject position and after-preposition position has increased these years with a significance level of p < 0.1. As for sociolinguistic factors involved, both gender difference and age difference were found. College males and high school males preferred to speak conservatively than our female subjects did. High school males and high school female used more disposal constructions than the other senior subjects did. The research questions of the present study are answered. Nevertheless, further study is absolutely needed. The instrument of the present study is Chinese-English and English-Chinese translation tests, and the sample numbers are quite restricted. Subjects’ performances may still be influenced by the written language. Even though to double check whether this is the case or not, we have also made a careful examination of a database corpus of transcripts from a popular TV show, which is not spontaneous discourse data. Thus, study based on spontaneous discourse database involving a sufficient of speakers is needed before a firm conclusion can be drawn.