In　this　paper,　we　investigate　the　min-max　correlation　clustering　problem　with　outliers,　which　is　a　combination　of　the　min-max　correlation　clustering　problem　with　the　robust　clustering.　We　first　prove　that　the　problem　is　NP-hard　to　obtain　any　finite　approximation　algorithm.　Then　we　design　an　approximation　algorithm　based　on　LP-rounding　technique　and　receive　a　bi-criteria　guarantee.