Abstract: Automatic and accurate pulse repetition pattern reconstruction of radar reconnaissance signals is an important content that is currently adapted to the development trend of electronic reconnaissance big data and the demand of data intelligent processing. Aimed at radar signals with fixed, stagger and other conventional pulse repetition intervals (PRI), frequent item extension algorithm is an effective method in automatic reconstruction of PRI patterns, however, which cannot perfectly meet the requirements for processing multi-emitters interleaved pulse trains. Due to the design of large bandwidth and wide directivity in reconnaissance receivers, interleaved pulse trains of multiple emitters are quite common in radar reconnaissance data. In order to reconstruct PRI pattern of each radar emitter in the interleaved pulse train without deinterleaving, this paper introduced the IS measure and presented a new termination criterion on the basis of the frequent item extension method. The simulation results showed that the method had strong ability of anti missing and interferential pulse and certain adaptability to the number of emitters, thus realizing the effective reconstruction of PRI pattern.