Memories associated with drug use increase vulnerability to relapse in substance use disorder (SUD) and there are no pharmacotherapies for the prevention of relapse. Previously, we reported a promising finding that storage of memories associated with methamphetamine (METH), but not memories for fear or food reward, is vulnerable to disruption by actin depolymerization in the basolateral amygdala complex (BLC). However, actin is not a viable therapeutic target because of its numerous functions throughout the body. Here we report the discovery of a viable therapeutic target, nonmuscle myosin II (NMIIB), a molecular motor that supports memory by directly driving synaptic actin polymerization. A single intra-BLC treatment with Blebbistatin, a small molecule inhibitor of class II myosin isoforms, including NMIIB, produced a long-lasting disruption of context-induced drug seeking (at least 30 days). Further, post-consolidation genetic knockdown of Myh10, the heavy chain of the most highly expressed NMII in the BLC, was sufficient to produce METH-associated memory loss. Blebbistatin was found to be highly brain penetrant. A single systemic injection of the compound selectively disrupted the storage of METH-associated memory and reversed the accompanying increase in BLC spine density. This effect was specific to METH-associated memory, as it had no effect on an auditory fear memory. The effect was also independent of retrieval, as METH-associated memory was disrupted twenty-four hours after a single systemic injection of Blebbistatin delivered in the home cage. Together, these results argue for the further development of small molecule inhibitors of nonmuscle myosin II as potential therapeutics for the prevention of SUD relapse triggered by drug associations.