Pan-histone deacetylase (HDAC) inhibitors, which inhibit 11 HDAC isoforms, are widely used to induce Epstein-Barr virus (EBV) lytic cycle in EBV-associated cancers in vitro and in clinical trials. Here, we hypothesized that inhibition of one or several specific HDAC isoforms by selective HDAC inhibitors could potently induce EBV lytic cycle in EBV-associated malignancies such as nasopharyngeal carcinoma (NPC) and gastric carcinoma (GC). We found that inhibition of class I HDACs, particularly HDAC-1, -2 and -3, was sufficient to induce EBV lytic cycle in NPC and GC cells in vitro and in vivo. Among a panel of selective HDAC inhibitors, the FDA-approved HDAC inhibitor romidepsin was found to be the most potent lytic inducer, which could activate EBV lytic cycle at~0.5 to 5 nM (versus~800 nM achievable concentration in patients' plasma) in more than 75% of cells. Upregulation of p21 WAF1 , which is negatively regulated by class I HDACs, was observed before the induction of EBV lytic cycle.The upregulation of p21 WAF1 and induction of lytic cycle were abrogated by a specific inhibitor of PKC-d but not the inhibitors of PI3K, MEK, p38 MAPK, JNK or ATM pathways. Interestingly, inhibition of HDAC-1, 22 and 23 by romidepsin or shRNA knockdown could confer susceptibility of EBV-positive epithelial cells to the treatment with ganciclovir (GCV). In conclusion, we demonstrated that inhibition of class I HDACs by romidepsin could potently induce EBV lytic cycle and mediate enhanced cell death with GCV, suggesting potential application of romidepsin for the treatment of EBV-associated cancers.