The state of Florida is developing a risk-based decision support model, specifically for its state-maintained bridges, to handle the occurrence of natural hazards, such as hurricanes, tornadoes, floods, landslides, and wildfires. Many bridges in Florida are vulnerable to damage caused by these hazards. With a focus on hurricanes, this paper develops prediction models (estimates of likelihoods) based on historical data of hazard events in Florida, as well as on existing national models. The Poisson process was assumed for the occurrence of natural events; predictions were estimated for the annual probability of occurrence. It was observed that, on average, the probability of hurricane occurrence at bridge locations decreased with the hurricane intensity (category number). In comparison, wildfires, another common hazard, had the highest likelihood estimates of the natural hazards. With a focus on physical damage to bridges, the consequences of hurricanes were also estimated on the basis of historical data within Florida. The damaged bridge components and elements were identified, and the levels of damage and the associated repair costs were assessed.