During the 12 years since basin-wide, BMP program implementation in 1995, the Everglades Agricultural Area (EAA) basin has achieved an average phosphorus (P) load reduction of greater than 50% relative to the baseline period from 1978 to 1988.  The EAA has organic soils that are drained through a network of pumps and canals, and farms are monitored for their drainage water P concentration and flow. Although the implemented Best Management Practices (BMP) point system has been quite effective in reducing P loads, there is still uncertainty concerning the efficacy of BMPs as seen by the variability of reduction in the different sub-basins in the EAA. Factors that may be affecting BMP performance include rainfall, soil depth, land use, farm location within the EAA, irrigation water quality, farm size, and rainfall detention amount. The objective of this work was to conduct a detailed statistical evaluation using multiple regression, principal component, and classification and regression tree analyses to evaluate the factors affecting P loads.  A database of ten farms monitored in the EAA for ten years for canal levels, rainfall, discharge flows, discharge water P concentrations, BMPs implemented, and crop land use were used.  Statistical analysis results indicate that in general P loading is associated with water management practices. However, there are other factors that may be impacting P loading including irrigation water quality.