The purpose of this note is to project private pension funds to 1990. Private pension fund assets totaled $254 billion in 1976. These assets were almost exclusively managed by financial institutions. Life insurers alone managed more than $ 100 billion in private pension funds and their growth is of interest to financial institutions, especially life insurers. Private pension fund assets are also of interest to financial institutions because of their role in the capital markets. For example, in 1974 private pension funds supplied $16.8 billion of the $176.2 billion advanced to the non-financial sectors in the credit market. Pension fund assets also accounted for over 30 percent of all stockholders' equity [4, pp. 15-19]. Private pension funds directly affect economic growth and stability, which have an impact on the return to financial institutions. Private pension funds are the largest single category of private saving in the United States. The savings function is related to economic growth because it makes capital formation possible. In terms of stability, if actual savings does not equal desired investment, output and prices will fluctuate and the reliability of economic forecasts will decrease. This economic instability increases the risk of many financial investments. The remainder of this study is an attempt to forecast private pension fund assets to 1990. The model is based upon the assumption that private pensions constitute a system with dynamics that will persist over the next decade and whose essence can be captured over another 10 to 15 years. The model developed in this study, like many business forecasting models, requires projections of other economic factors for use as independent variables. In building the model, an effort was made to select independent variables that were either known or whose forecasts, due to either economic or institutional factors, are relatively reliable when compared to alternative variables. Although it is recognized that the use of economic projections, especially long-run forecasts, should be avoided whenever possible, the nature of the model necessitates them. Data Resources Incorporated (DRI) forecasts are used in this study, unless otherwise noted, because they are highly respected