This article considers the issue of technological changes in a replacement cost accounting system. When such changes occur and the firm continues to utilize the "old" asset, should the replacement cost be measured by reference to the current cost of the old asset actually in use? Or, alternatively, should the replacement cost reflect the cost of the technologically improved asset that is not now being utilized by the firm? The analysis indicates that the two approaches are economically equivalent in many circumstances. However, while their information content may be identical, various measurement problems exist which tend to make one method preferable to the other under certain conditions. These issues are explored in various market settings.