Phosphorus (P) in lake water is commonly partitioned into particulate P and dissolved P by membrane filtration, and dissolved P is then fractionated into soluble reactive P (SRP, reactive with molybdate) and dissolved unreactive or organic P (DOP). Much of what is known about DOP is derived from radiotracer studies using gel chromatography, and summarized by a kinetic model (Lean, Science 179:678-680, 1973a; Lean, J Fish Res Board Can 30:1525-1536, 1973b. Since this work, several relevant discoveries have been made on the role of enzymes, viruses and zooplankton in regenerating dissolved P, and the role of filtration damage in generating dissolved P in filtrates. Herein we present the results of new radiotracer experiments on the fate of DOP in lake water filtrates, consistent with the hypothesis that some of the high molecular weight organic P breaks down spontaneously to smaller molecules, which in turn break down to PO 4 3− . We use inhibitors, including competitive inhibitors of phosphatases and a commercial product (RNA-later ® ) to support the hypothesis that the larger molecules include nucleic acids, and that the smaller molecules are substrates for alkaline phosphatase. We also find that colloidal P (i.e., P > 5000 MW according to gel filtration) includes some virus-sized material that can be collected on 0.02 or 0.03 µm filters. Finally, we provide a new model of the cycling of epilimnetic P that is consistent with these and earlier observations.