The oak longhorned beetle (OLB),
Massicus raddei
(Blessig, 1872) (Coleoptera: Cerambycidae), is widely distributed in Asia (China, the Korean Peninsula, Japan, Vietnam and the Russian Far‐East), but pest outbreaks have occurred only in Liaoning Province and Jilin Province of China. In order to explore possible mechanisms of local population outbreaks and characterize the genetic diversity and genetic structure of
M. raddei
across its range in China, three mitochondrial genes (
COI, Cytb,
and
COII
) were sequenced and analyzed for seven
M. raddei
populations collected from six provinces in China. From these different populations, we found a high haplotype and nucleotide diversity. Haplotype networks and phylogenetic analyses both demonstrate apparent genetic diversification between SC (southern China) and NC (northern China) population groups. A set of 21 pairwise comparisons for
Fst
(pairwise fixation indices) and
Nm
(genetic flow index) showed significant genetic differentiation and limited gene flow except for two pairs, Shandong (SD) and Liaoning (LN), and Anhui (AH) and Henan (HN). This pattern suggested that the periodic outbreak of the LN population could not be attributed to the absence of genetic flow with other spatial populations and that regional environmental factors might be responsible. AMOVA (Analysis of molecular variance) showed that the greater molecular genetic variation was among populations. Based on Tajima's
D
statistic, Fu's
Fs
, and the mismatch distribution test, we determined that the seven populations sampled were stable and had not experienced any recent population expansion. The fact that all the sampled populations showed only unique haplotypes and lacked shared or ancestral haplotypes, as well as the nonstar‐like distribution of haplotype network for concatenated genes, collectively provided powerful evidence of the stable and isolated nature of most populations. The high genetic differentiation and spatial genetic structuring among populations are both likely related to the beetle's moderate flight capacity, regional variation in host tree species and microclimate, as well as the geographic distance between sampling sites.