ABSTRACTThis paper discusses the Stochastic Dominance (SD) approach to the evaluation of risky assets. Given a set of portfolios, the familiar EV procedure chooses a set of EV‐efficient portfolios while the SD procedures choose SD‐efficient sets that usually are quite similar to (but not identical with) the EV‐efficient set. The SD approach can be employed as a refinement of the EV model or as an alternative method of evaluating portfolios. The SD view has certain conceptual advantages in the screening of a set of portfolios, but the EV model has the important advantage of an optimizing algorithm that builds efficient portfolios. Also, the SD approach requires more data than the EV app