Although many formalisms for reasoning about action and change have been proposed in the literature, their semantic adequacy has primarily been tested using tiny domains that highlight some particular aspect or problem. However, since some of the classical problems are completely or partially solved and since powerful tools are available, it is now necessary to start modeling more complex domains. This paper presents a methodology for handling such domains in a systematic manner using an object-oriented framework and provides several examples of the elaboration tolerance exhibited by the resulting models.