Despite having limitless access to knowledge during this era of information explosion, we all acknowledge that big data is not knowledge. Indeed, humans are endowed with tremendous abilities to reason and infer the hidden relational structure behind fragmented outside events, an essential index for intelligence. In this talk, I will present my lab’s recent works aiming to understand how relational structure knowledge, from simple sequence to hierarchy and network, is represented, maintained, and learned in the human brain. Our findings support dissociated structure-content neural representations and an essential role of relational structure in mediating memory and learning.