We show that rationalizability in a variety of general preference models yields the unique outcome of iterated strict dominance. The result implies that rationalizable strategic behavior in these preference models is observationally indistinguishable from that in the subjective expected utility model. Our indistinguishability result can be applied not only to mixed extensions of finite games, but also to other important applications in economics, for example, the Cournot-oligopoly model. JEL Classification: C70, D81.