Antonyms co-occur far more frequently than we might expect to by chance alone. However, few studies have adopted a corpus-based approach to examine antonym co-occurrences in Chinese, and disyllabic antonyms in Chinese have long been ignored in the literature. Therefore, this study turns to the Chinese Gigaword Corpus and includes both monosyllabic and disyllabic pairs for analysis, aiming to identify contrastive constructions in which antonyms in Chinese often co-occur and to examine antonym co-occurrence sequences in Chinese. First, it is found that most functional classes of antonymy in Chinese are associated with specific constructions. Second, a positive antonym in Chinese (e.g., dui ‘right’) usually precedes its negative (e.g., cuo ‘wrong’) counterpart when they co-occur, and the notion of positivity needs to be defined in a broad sense. Third, the behaviour of an antonym pair in Chinese can be influenced by its morphosyllabic structure and some language-external factors such as cognitive functions and socio-cultural values. Our understanding of antonymy in Chinese has been updated, and some striking similarities between Chinese and English in the use of antonyms have also been revealed. It is hoped that the findings of this study will invite more attention to the syntagmatic dimension of antonymy in Chinese.