A new expression for ion leakage from plant tissue, the tissue ionic conductance (g(Ti)), is compared with electrical conductivity (EC) and a commonly used damage index (I(d)) to test the ability of each expression to correctly describe leakiness in two model systems representing examples of physiological processes with well-known effects on membrane permeability. In experiments in which drought-acclimated leaves were compared with nonacclimated leaves and senescing leaves were compared with nonsenescing leaves, I(d) contradicted our expectation that acclimated tissue would be less leaky than nonacclimated tissue, and g(Ti) and EC confirmed this expectation. In a comparison of senescing and nonsenescing tissue, I(d) again contradicted our expectation that senescing tissue would be more leaky than nonsenescing, and EC and g(Ti) were confirming. Using a diffusion analysis approach, we show that I(d) fails to account for variation in the concentration gradient between the tissue and the bathing solution and variation in the surface area through which efflux occurs. Furthermore, because I(d) is a parameter that relates treatment performance to control performance as a percentage value, it distorts the actual differences among treatments. The resulting artifacts lead to a presentation of membrane integrity which is probably incorrect. EC is a more direct measurement of net ion efflux and appears to be less vulnerable to artifact. However, because g(Ti) is the only expression that explicitly includes chemical driving force and tissue surface area, it is the most reliable of the three expressions.