The liquid metal protection concept is proposed mainly to prolong the lifetime of a divertor or a limiter in a fusion reactor. This attractive idea for protection requires studying a wide range of problems associated with the use of liquid-metals in fusion reactors. In this work the protection by liquid-metals has concentrated on predictions of the loss rate of the film to the plasma, the operating surface temperatures required for the film, and the potential tritium inventory requirement. The effect of plasma disruptions on the liquid metal film is also evaluated. Other problems such as liquid metal compatibility with structural materials, magnetic field effects, and the effect of liquid metal contamination on plasma performance are discussed. Three candidate liquid-metals are evaluated, i.e., lithium, gallium, and tin. A wide range of reactor operating conditions valid for both near term machines (INTOR and ITER) and for the next generation commercial reactors (TPSS) are considered. This study has indicated that the evaporation rate for candidate liquid metals can be kept below the sputtering rate for reasonable operating temperatures and plasma edge conditions. At higher temperatures, evaporation dominates the losses. Impurity transport calculations indicate that impurities from the plate should not reach the main plasma. One or two millimeters of liquid films can protect the structure from severe plasma disruptions. Depending on the design of the liquid metal protection system, the tritium inventory in the liquid film is predicted to be on the order of a few grams.