A series of redox flow batteries utilizing mixed addenda (vanadium and tungsten), phosphorus-based polyoxometalates (A-α-PV3W9O40(6-), B-α-PV3W9O40(6-), and P2V3W15O62(9-)) were prepared and tested. Cyclic voltammetry and bulk electrolysis experiments on the Keggin compounds (A-α-PV3W9O40(6-) and B-α-PV3W9O40(6-)) established that the vanadium centers of these compounds could be used as the positive electrode (PV(IV)3W(VI)9O40(9-)/PV(V)3W(VI)9O40(6-)), and the tungsten centers could be used as the negative electrode (PV(IV)3W(VI)9O40(9-)/PV(IV)3W(V)3W(VI)6O40(12-)) since these electrochemical processes are separated by about 1 V. The results showed that A-α-PV3W9O40(6-) (where A indicates adjacent, corner-sharing vanadium atoms) had coulombic efficiencies (charge in divided by charge out) above 80%, while the coulombic efficiency of B-α-PV3W9O40(6-) (where B indicates adjacent edge-sharing vanadium atoms) fluctuated between 50% and 70% during cycling. The electrochemical yield, a measurement of the actual charge or discharge observed in comparison with the theoretical charge, was between 40% and 50% for A-α-PV3W9O40(6-), and (31)P NMR showed small amounts of PV2W10O40(5-) and PVW11O40(4-) formed with cycling. The electrochemical yield for B-α-PV3W9O40(6-) decreased from 90% to around 60% due to precipitation of the compound on the electrode, but there were no decomposition products detected in the solution by (31)P NMR, and infrared data on the electrode suggested that the cluster remained intact. Testing of P2V3W15O62(9-) (Wells-Dawson structure) suggested higher charge density clusters were not as suitable as the Keggin structures for a redox flow battery due to the poor stability and inaccessibility of the highly reduced materials.