Identifying overlapping protein complexes in protein-protein interaction (PPI) networks can provide insight into cellular functional organization and thus elucidate underlying cellular mechanisms. Recently, various algorithms for protein complexes detection have been developed for PPI networks. However, majority of algorithms primarily depend on network topological feature and/or gene expression profile, failing to consider the inherent biological meanings between protein pairs. In this paper, we propose a novel method to detect protein complexes using pseudo-clique extension based on fuzzy relation (PCE-FR). Our algorithm operates in three stages: it first forms the nonoverlapping protein substructure based on fuzzy relation and then expands each substructure by adding neighbor proteins to maximize the cohesive score. Finally, highly overlapped candidate protein complexes are merged to form the final protein complex set. Particularly, our algorithm employs the biological significance hidden in protein pairs to construct edge weight for protein interaction networks. The experiment results show that our method can not only outperform classical algorithms such as CFinder, ClusterONE, CMC, RRW, HC-PIN, and ProRank +, but also achieve ideal overall performance in most of the yeast PPI datasets in terms of composite score consisting of precision, accuracy, and separation. We further apply our method to a human PPI network from the HPRD dataset and demonstrate it is very effective in detecting protein complexes compared to other algorithms.