Studies of polymer electrolyte solutions for lithium-polymer batteries are described. Two different salts, lithium bis(trifluoromethane-sulfonyl)imide (LiTFSI) and lithium trifluoromethanesulfonate (LiTf), were dissolved in a variety of polymers. The structures were all based upon the ethylene oxide unit for lithium ion solvation, and both linear and comb-branch polymer architectures have been examined. Conductivity, salt diffusion coefficient and transference number measurements demonstrate the superior transport properties of the LiTFSI salt over LiTf. Data obtained on all of these polymers combined with LiTFSI salts suggest that there is a limit to the conductivity achievable at room temperature, at least for hosts containing ethylene oxide units. The apparent conductivity limit is 5 X 10(-5) S/cm at 25 degrees C. Providing that the polymer chain segment containing the ethylene oxide units is at least 5-6 units long, there appears to be little influence of the polymer framework to which the solvating groups are attached. To provide adequate separator function, the mechanical properties may be disconnected from the transport properties by selection of an appropriate architecture combined with an adequately long ethylene oxide chain. For both bulk and interfacial, transport of the lithium ions, conductivity data alone is insufficient to understand the processes that occur. Lithium ion transference numbers and salt diffusion coefficients also play a major role in the observed behavior and the transport properties of these polymer electrolyte solutions appear to be quite inadequate for ambient temperature performance. At present, this restricts the use of such systems to high temperature applications. Several suggestions are given to overcome these obstacles.