This paper presents an application of Sandewall's methodology for assessing nonmonotonic entailment criteria for reasoning about actions and change. We establish the correctness of Reiter's general solution to the frame problem for a broad and well characterized class of problems. This is done by: (1) identifying a nonmonotonic entailment criterion corresponding to Reiter's solution; (2) discovering ontological and epistemological assumptions that we believe underly the domains of possible applicability of Reiter's approach; (3) defining an underlying semantics corresponding to these assumptions; and (4) providing a semantic correspondence between the models selected by the entailment criterion and the models obtained from the underlying semantics. In particular, we assess the applicability of Reiter's solution to formalize database updates. We also compare our methodology to Kartha's evaluation of Reiter's general solution with respect to action logic semantics. Many solutions to the frame problem have been proposed in the literature, some are monotonic and some are nonmonotonic. Reiter recently proposed a solution and its application to the domain of database update specification, but no proof of soundness or completeness has been provided. We assess Reiter's monotonic solution to the frame problem in the database domain following Sandewall's methodology and we provide a more general underlying semantics that will serve as basis for the assessment of other solutions as well.