Antibiotic resistance in Mycobacterium tuberculosis (Mtb) is a major public health problem worldwide. Therefore, it is of great significance to study the mutational pathways of susceptive Mtb to drug resistance. Multi-omics approach is ideal to explore more about other molecular mechanisms implicated in drug resistance. In this study, we respectively constructed continuous capreomycin (CAP) and amikacin (AKM)-resistant strain models in vitro using the increasing drug concentration gradients. Based on these models, we use multi-omics techniques and biological experiments to reveal the drug resistance mutations caused by CAP or AKM, and obtain other factors that can cause drug resistance besides the mutations, and explain the differences in cross-resistance pathways caused by AKM and CAP. Our results showed that in addition to Gly232AsptlyA and Trp120fstlyA directly associated with low level resistance to CAP, some mutations, including Ala48ValmmaA2, Gln19ArgrpmA and eis (c.-14c>t), were associated with cross-resistance caused by CAP. The transcriptome-proteome-lipid metabolome analysis of representative induced strains showed that cross-resistance of CAP-induced strains was caused by a combination of many aspects, which was obviously different from the induction of AKM. The results of this study will advance our understanding of the CAP-resistant mechanisms. To a certain extent, it is also conducive to improve the clinical understanding of capreomycin and rational use of CAP.