By focusing attention on managers' concept of risk and the methods they use to measure risk, this paper attempts to lessen the disparity between current theory and practice by suggesting a method that will make IRR and NPV more generally applicable. We attempt to do this through a methodology that combines some PERT assumptions with traditionally accepted academic techniques and those used by practicing business managers to provide a probability distribution of the cumulative net present value (CPV). From these data we develop a probability distribution of the discounted payback period.