Two experiments tested the role of morphemic information and interword spacing in reading in experienced and inexperienced Chinese readers. Chinese is normally written in "hanzi," or characters, which mostly represent monosyllabic morphemes, but it can also be written in "pinyin," or romanised Chinese, which represents phonemes and is word-spaced. While previous research has shown that Chinese readers are slower with pinyin than hanzi materials, this has mostly been explained in terms of lack of proficiency in pinyin reading. The present study aimed at testing whether pinyin reading may be slow because morphemic information is needed for fluent Chinese reading, and phonemic information alone is not sufficient; for this purpose, the study included not only adults but also primary school students, who are experienced pinyin readers and unproficient hanzi readers. Participants performed a sentence-picture verification task. Sentences were written with morphemic or phonemic information (in hanzi or pinyin, respectively), and with interword or inter-morpheme spacing. Removing morphemic information had strong negative effects on all readers, including children. Adding interword spacing had no facilitative effects, and had some negative effects, especially with children. Results reveal the important role of morphemic information in Chinese reading, and fail to support the universality of the facilitative effects of interword spacing.