Antimicrobial resistance (AMR) stands as an escalating public health crisis fueled by antimicrobial residues in the environment, particularly in soil, which acts as a reservoir for antimicrobial resistance genes (ARGs). Merely quantifying the total extractable concentration of antimicrobials, instead of bioavailable fractions, may substantially underestimate their minimal selection concentration for propagating ARGs. To shed light on the role of bioavailability in ARG abundance within soil, a systematic bioavailability assessment method was established for accurately quantifying the partitioning of multi-class antimicrobials in representative Chinese soils. Microcosm studies unveiled that antimicrobials persisting in the bioavailable fraction could potentially prolong their selection pressure duration to trigger AMR. Notably, the co-occurrence of pesticide or steroid hormone influenced the development trends of ARG subtypes, with fluoroquinolone resistance genes (RGs) being particularly susceptible. Partial least squares path model (PLS-PM) analysis uncovered potentially distinct induction mechanisms of antimicrobials: observable results suggested that extractable residual concentration may exert a direct selection pressure on the development of ARGs, while bioavailable concentration could potentially play a stepwise role in affecting the abundance of mobile genetic elements and initiating ARG dissemination. Such unprecedented scrutinization of the interplay between bioavailable antimicrobials in soils and ARG abundance provides valuable insights into strategizing regulatory policy or guidelines for soil remediation.