Fine-Grained integrity check for forensic data becomes an important demand of computer forensics. It will mitigate the disasterous effect on the data by some random errors or the intentional forging modification.Unfortunately, the traditional method generates a hash for every piece of small data and produces a large amount ofhash data. These hash data are random data and can not be compressed in a normal way. It has a great negativeimpact on storing hash data and transmitting them over network. Based on the error correction coding theory, afine-grained integrity check method, an integrity indication code, is proposed. The properties of the integrity indication code are analyzed. Combinatorial codes for one error in a group of data objects are also proposed. Hash data can be compressed hundredfold using combinatorial codes. This paper provides a fundamental support for further research on fine-grain data integrity check method and related applications.