In information-centric networks (ICN), each node is equipped with a cache and it can improve content access and transmission efficiency. However, the built-in cache capacity is small, and it can’t completely store huge amounts of content transmitted. The existing caching schemes lack choices in content placement and balanced distribution, thus, leading to the problem of low cache hit ratio and high user access time delay. To solve the problem, this paper proposes a popularity and gain based caching scheme (PGBCS). It refines the content object from the content file to the chunk, thus achieves fine-grained cache management. The strategy makes use of the feature of content chunk popularity, and takes the factors that affect the overall caching gain into consideration, and realizes the placement and replacement of content chunks through comprehensive measure. The simulation results show that compared with other schemes, this method can effectively promote node’s cache hit ratio, reduce user request delay and further raise network service quality.