Summary: In this paper we show that, given a monotone preference order that satisfies weak independence, the expected utility model (as a \textit{weak} or \textit{strong} representation) can be directly obtained from intuitive conditions. In particular, a weak expected utility representation can be constructed by using \textit{substitutability}, while the stronger counterpart can be derived by using \textit{solvability}. We provide simple examples which demonstrate that our characterization of the expected utility preferences is tight. Unlike many other expected utility characterizations, our proofs are constructive making them especially useful when testing the model.