Acute aortic dissection (AAD) is a common and fatal disease for which pathogenesis is largely unknown. Recently we and others reported that JAK/STAT-activating cytokines are highly expressed in human and mouse models of AAD. We also reported in this meeting last year that STAT3 is activated both in macrophages and in vascular smooth muscle cells (VSMCs) during AAD development, and STAT3 activation in macrophages promotes AAD progression. However, the role of STAT3 in VSMCs is unknown in the context of AAD. To address this question, we used a mouse model of AAD that was induced by continuous infusion of beta-aminopropionitrile, an inhibitor of lysyl oxidase that cross-links collagen and elastin, and angiotensin II (BAPN+AngII). BAPN+AngII caused thoracic and suprarenal AAD starting at 7 days of the administration, and most of the mice developed AAD within 14 days. We first performed the transcriptome analysis 3 days after the BAPN+AngII infusion, when aorta showed no obvious structural changes. The induced genes by BAPN+AngII were enriched for those encoding proinflammatory cytokines. The suppressed genes by BAPN+AngII were enriched for those encoding extracellular matrix proteins that are normally maintained by VSMCs and cytoskeletons of VSMCs, indicating that profound changes in VSMCs function precedes AAD development. To investigate into the function of STAT3 in VSMCs, we used the smooth muscle-specific knockout mice for STAT3 (smSTAT3-KO). We also used the smooth muscle-specific knockout mice for SOCS3, a negative feedback regulator of STAT3, to enhance the STAT3 signal specifically in smooth muscle cells (smSOCS3-KO). We determined the AAD severity by the length of the lesions with enlarged aortic diameter. BAPN+AngII caused more severe AAD in smSTAT3-KO mice compared with wild type mice in the descending aorta (WT; 0.069±0.024 vs. smSTAT3; 0.214±0.084 mm/g body weight, P=0.028). In contrast, smSOCS3-KO mice developed less severe AAD in the aortic arch (WT; 0.093±0.018 vs. smSOCS3-KO; 0.018±0.013 mm/g body weight, P=0.014). These results indicate that dysfunction of VSMCs with proinflammatory response precedes the AAD development, and STAT3 activation in VSMCs protects aorta from AAD, possibly by maintaining the VSMC function.