Biscayne National Park, located in Biscayne Bay in southeast Florida, is one of the largest marine parks in the country and sustains a large natural marine fishery where numerous threatened and endangered species reproduce. In recent years, the bay has experienced hypersaline conditions (salinity greater than 35 practical salinity units) of increasing magnitude and duration. Hypersalinity events were particularly pronounced during April to August 2004 in nearshore areas along the southern and middle parts of the bay. Prolonged hypersaline conditions can cause degradation of water quality and permanent damage to, or loss of, brackish nursery habitats for multiple species of fish and crustaceans as well as damage to certain types of seagrasses that are not tolerant of extreme changes in salinity. To evaluate the factors that contribute to hypersalinity events and to test the effects of possible changes in precipitation patterns and canal flows into Biscayne Bay on salinity in the bay, the U.S. Geological Survey constructed a coupled surface-water/groundwater numerical flow model. The model is designed to account for freshwater flows into Biscayne Bay through the canal system, leakage of salty bay water into the underlying Biscayne aquifer, discharge of fresh and salty groundwater from the Biscayne aquifer into the bay, direct effects of precipitation on bay salinity, indirect effects of precipitation on recharge to the Biscayne aquifer, direct effects of evapotranspiration (ET) on bay salinity, indirect effects of ET on recharge to the Biscayne aquifer, and maintenance of mass balance of both water and solute. The model was constructed using the Flow and Transport in a Linked Overland/Aquifer Density Dependent System (FTLOADDS) simulator, version 3.3, which couples the two-dimensional, surface-water flow and solute-transport simulator SWIFT2D with the density-dependent, groundwater flow an solute-transport simulator SEAWAT. The model was calibrated by a trial-and-error method to fit observed groundwater heads, estimated base flow, and measured bay salinity and temperatures from 1996 to 2004, as well as the location of the freshwater-saltwater interface in the aquifer, by adjusting ET rate limiters, canal vertical hydraulic conductance, leakage rate coefficients (transition-layer thickness and hydraulic conductivity), Manning's n value, and delineation of rainfall zones. Although flow budget calculations indicate that precipitation, ET, and groundwater flux into the bay represent a small portion of the overall budget, these factors may be important in controlling salinity in some parts of the bay, for example the southern parts of the bay where the canal system is not extensively developed or controlled. The balance of precipitation and ET during the wet season generally results in a reduction of bay salinity, whereas the balance of precipitation and ET during the dry season generally results in an increase in bay salinity. During years when wet season precipitation is lower than average, for example less than 70 percent total precipitation for an average year, ET could outweigh precipitation over the bay for essentially the entire year. Hypersaline conditions are prone to occur near the end of the dry season because precipitation rates are generally lower, canal discharge rates (which are strongly correlated to precipitation rates) are also generally lower, and ET rates are higher than during the rest of the year. The hypersalinity event of 2004 followed several years of relatively low precipitation and correspondingly reduced canal structure releases and was unusually extensive, continuing into July. Thus, hypersalinity is ultimately the result of a cumulative deficit of precipitation. The model was used to test the effects of possible changes in canal flux and precipitation. Simulation results showed that by increasing, reducing, or modifying canal discharge rates, the effects on salinity in the bay were more pronounced in the northern part of the bay where there are more canals and canal-control structures. By doubling and halving precipitation, the effects on bay salinity were more pronounced in the southern part of the bay than in the northern part of the bay where there are fewer canals and canal-control structures. The model is designed to quantify factors that contribute to hypersaline conditions in Biscayne Bay and may be less appropriate for addressing other issues or examining conditions substantially different from those described in this report. Model results must be interpreted in light of model limitations, which include representation of the system and conceptual model, uncertainty in physical properties used to describe the system or processes, the scale and discretization of the system, and representation of the boundary conditions.