The term ''stability,'' in its ecological context, has been applied to populations, communities, and, most recently, to ecosystems with such an array of meanings that it has almost become a panchreston. This review of selected literature points out the problems inherent in relying on either static measures (e.g., species diversity, connectivity, or energy balance) as a measure of complexity to arbitrarily rank the relative stabilities of ecosystems, or the dynamic approach of analyzing the mathematical stability of linear or nonlinear models. The research direction emphasized is that of development of new measures of ecosystem complexity based on ecosystem level characteristics (e.g., nutrient retention, or ecosystem metabolism) and relating these measures of complexity to an a priori defined ecosystem stability. (ERA citation 02:011162)