This　paper　proposes　a　digital　twin　(DT)　modeling　method　for　the　lifecycle　state　evaluation　of　cable-stayed　bridges,　which　cover　a　large　time　scale　and　simultaneously　involve　macroscopic　and　microcosmic　exploration.　First,　the　DT　model　at　the　design　stage　corresponds　to　the　computer-aided　design　(CAD)　model　and　then　evolves　to　the　computer-aided　engineering　(CAE)　model　during　the　construction　stage.　After　that,　the　DT　model　for　the　operational　stage　starts　from　the　completion　model　of　the　bridge　and　keeps　evolving　with　the　bridge　physical　entity　in　the　long-term　service.　An　information　interaction　media　is　used　for　the　information　interchange　between　the　DT　model　and　the　bridge　physical　entity.　And　a　fidelity　index　gives　the　quantitative　estimation　of　similarity　between　the　DT　model　and　the　bridge　physical　entity.　Finally,　the　proposed　method　has　been　verified　against　a　cable-stayed　bridge　model,　whose　DT　model　could　reflect　the　actual　state　evolution　of　the　physical　entity.