Three experiments were conducted to investigate how syntactic-category and semantic information is processed in visual word recognition. The stimuli were two-character Chinese words in which semantic and syntactic-category ambiguities were factorially manipulated. A lexical decision task was employed in Experiment 1, whereas a semantic relatedness judgment task and a syntactic-category judgment task were adopted in Experiments 2 and 3, respectively. A semantic ambiguity disadvantage was observed in all three experiments, whereas a syntactic-category ambiguity disadvantage was only found in Experiment 3. In addition, no significant semantic ambiguity x syntactic-category ambiguity interaction was obtained across the three experiments. These results are consistent with the view that in isolated visual word recognition, the semantic information is crucial, but the syntactic-category information is not, at least in Chinese. (Contains 2 figures, 6 footnotes and 4 tables.)