In　most　real-world　cases,　an　in-service　structure　doesn＇t　always　follow　the　two-state　hypothesis　under　which　the　structure　stays　at　an　intact　or　completely　failed　state.　Structural　failure　is　sometimes　regarded　as　a　fuzzy　event,　and　the　actual　failure　boundary　has　a　certain　level　of　ambiguity　that　affects　the　structural　limit　state　function　under　a　fuzzy　failure　criterion.　Under　such　circumstance,　structural　reliability　should　be　solved　within　a　hybrid　reliability　analysis　framework　involving　the　coupled　effect　of　randomness　and　fuzziness.　A　fuzzy　Bayesian　interval　estimation　strategy　has　been　proposed　for　this　purpose.　Structural　parameters　and　external　loads　having　fuzziness　are　decomposed　and　extended　to　fuzzy　sets.　The　interval　bounds　of　the　distribution　characteristics　of　the　fuzzy　parameters　and　loads　are　estimated　using　the　fuzzy　Bayesian　estimation.　Then　an　equivalent　performance　function　is　defined　considering　the　fuzziness　of　the　failure　criterion.　After　that,　the　failure　probability　is　computed　under　different　interval　combinations.　The　structural　failure　probability　is　expressed　by　an　interval,　instead　of　a　traditional　deterministic　value.　The　solution　process　provides　a　better　estimation　of　failure　boundaries　taking　into　account　parameter　ambiguities.　The　proposed　method　has　been　successfully　verified　against　a　plane　steel　frame　structure　and　the　IASC-ASCE　benchmark　test　frame.　It　was　found　that　the　estimated　failure　probability　intervals　embraced　the　deterministic　value　predicted　by　the　Monte　Carlo　simulation.