The paper presents a decision support system (DSS) for managing and using recreational beaches. The DSS consists of (1) a sensor-assisted water quality monitoring system, (2) a multiple linear regression (MLR) model, developed with the Virtual Beach (VB) program, for predicting enterococci levels, and (3) a web-enabled Geographic Information System (GIS) platform for displaying beach water quality conditions. The MLR model was tested using a total of 945 sets of environmental and bacteriological data collected over 6 swimming seasons. It is found that the MLR model is capable of correctly predicting over 88% of beach advisories and over 80% of no advisory events. The web-enabled GIS platform has nowcasting and forecasting functions. The nowcasting function reduces the decision-making time from current 2–3 days to near real-time while the forecasting function changes the decision-making time from current 2–3 days behind to 1–3 days ahead of actual use of beaches, greatly improving the decision-making in beach management and reducing potential health risks of fecal pollution to beachgoers. While the DSS was specifically developed for the Holly Beach, USA, the methods used in this paper are generally applicable to other coastal beaches.