Considering　the　characteristics　of　urban　space　layout,　earthquake　and　its　secondary　disasters,　and　the　problems　to　restriction　and　nonlinearity　of　evacuation　capacity　evaluation,　the　factors　such　as　population,　buildings,　green　spaces,　roads,　seismic　refuges　for　evacuation,　major　hazard　installations　are　selected　as　evaluation　indexes　in　aspects　of　people,　objects,　environment　and　disaster,　based　on　the　principles　of　security,　reachability　and　evacuation　validity.　The　self-determined　intelligent　evaluation　method　of　urban　earthquake　evacuation　capacity　is　proposed　based　on　the　concept　of　dominance　and　intelligent　algorithms.　Firstly,　the　prior　order　of　the　evaluation　indexes　is　determined　by　calculating　the　absolute　dominance　degree　and　relative　dominance　degree.　Secondly,　the　position　weighted　vector　and　the　evaluation　value　vector　of　the　evaluation　object　are　solved　by　leapfrog　algorithm　optimizing　IOWA　operator.　Finally,　the　optimal　evaluation　vector　is　obtained　by　the　aggregated　model　of　evacuation　values.　Example　shows　that　the　evaluation　results　reflect　the　reality　of　the　urban　earthquake　evacuation　affected　by　a　variety　of　factors.　It　is　highly　acceptable,　and　provides　the　a　basis　for　decision　making　for　urban　earthquake　evacuation　planning.