The　gust　factor　commonly　is　used　in　wind　engineering　community　to　convert　mean　wind　speeds　into　gusty　winds,　which　exhibit　significant　variability　in　real　observations　of　typhoon　winds.　This　study　proposes　a　probabilistic　gust　factor　model　that　accounts　for　uncertainties　of　wind　speed　statistics.　The　statistical　characteristics　of　wind　speed　from　nine　typhoons,　including　the　mean,　standard　deviation,　skewness,　kurtosis,　power　spectral　density　(PSD)　parameter,　peak　factor,　and　gust　factor,　were　examined.　The　effects　of　nonstationary　characteristics　in　terms　of　time-varying　mean　wind　speed　and　non-Gaussian　attributes　of　fluctuating　winds　on　the　gust　factor　are　discussed.　These　wind　speed　statistics　were　incorporated　into　a　non-Gaussian　moment-based　translation　model　to　perform　the　Monte　Carlo　simulation　of　peak　factor　and　gust　factor.　The　simulation　results　for　different　gust　durations　were　juxtaposed　with　observations　to　substantiate　the　accuracy　of　the　probabilistic　model.　Subsequently,　a　standardization　framework　for　estimating　site-specific　probabilistic　gust　factor　curves　was　developed.　This　approach　was　applied　to　determine　the　gust　typhoon　wind　speed　hazard　curve　at　a　real　bridge　site　with　flat　open　terrain.　The　present　model　enables　the　consideration　of　gust　factor　dispersion　to　achieve　a　probabilistic　gust　wind　hazard　curve　and　facilitate　the　development　of　performance-based　wind　engineering.