Applying　acoustic　emission　method　to　understand　the　damage　and　failure　mechanism　of　hybrid　composites　is　challenge　for　industry.　For　this　study,　nine　types　of　carbon/glass　fiber-reinforced　hybrid　laminate　composite　specimens　with　different　structure　and　indentation　were　subjected　to　tensile　experiments.　The　AE　signals　collected　during　the　tensile　process　were　post-processed　using　cluster　analysis　based　on　Fuzzy　C-Means　algorithm.　It　was　found　that　the　AE　signals　can　be　divided　into　three　types　which　correspond　to　three　damage　modes,　and　the　AE　peak　frequency　characteristics　of　each　damage　mode　were　found　for　different　specimens.　Microcomputed　tomography　imaging　was　obtained　for　specimens　with　indention　before　loading　and　after　failure.　There　is　seen　to　be　a　good　correlation　between　the　damage　seen　with　the　Micro-CT　imaging,　the　mechanisms　identified　and　the　data　cluster　analyzed　by　using　AE　signals.　Results　prove　that　AE　signals　are　reliable　and　can　be　used　for　composite　structure　health　monitoring.