An empirical phosphorus (P) loading model modified from the original Vollenweider formulation has been used since the late 1970s to track progress toward a legally mandated P loading target for Lake Okeechobee, Florida. The loading target is designed to achieve, as an annual average for the pelagic region, a total P (TP) concentration of 40 μg L−1. This TP goal is not based on historic data from the lake, nor on how certain levels of TP might cause use-impairment or ecological harm. Nevertheless, our retrospective analyses indicate that the goal falls within a range of TP concentrations (26 to 92 μg L−1) derived from historical data, pelagic TP-algal bloom relationships, pelagic TP-chlorophyllarelationships, and a Florida lake regression model. When first applied to Lake Okeechobee, the modified Vollenweider model gave accurate estimates of pelagic TP, but now it under-predicts TP by nearly 50 percent. This may reflect time lags in lake responses to recent reductions in P loads, an increase in the relative magnitude of internal vs. external P loads, or a change of in-lake processing of P. The lake's P budget shows a decline over time in the net sedimentation of P. The Vollenweider model estimates P sedimentation based on a fixed empirical relationship using water residence time (τw), a parameter that has not displayed a significant historical trend. Given these issues, it is important to consider whether the existing model is an effective management tool for Lake Okeechobee. Our results indicate that the modified Vollenweider model may suffice as a coarse-scale tool for tracking progress in the eutrophication management program, with a major caveat: model predictions of pelagic TP at any given external loading rate may reflect what ispotentiallyattainable, if internal loading rates decrease to their previous lower levels. A more complex dynamic model is being developed, which accounts for sediment-water P exchanges. The new model should provide more accurate estimates of pelagic TP, as well as estimates of recovery time and predictions of short-term responses to management actions.