The Iowa-curve methodology is based on the use of survival curves and is commonly employed for estimating the useful life of industrial property subject to depreciation charges. This article is an overview of this methodology. We point out some of the problems associated with it, and suggest ways of rectifying them. When there is paucity of retirement data for a single group of properties, the methodology calls for the consolidation of properties from different groups. We point out that such consolidations can lead to the gross underestimation of useful life due to an apparent paradox, which is reminiscent of Simpson's paradox.