This paper concerns recent efforts to elucidate the microscopic physical mechanisms responsible for intermetallic compound formation and to understand thereby Miedema's empirical rule for compound formation. More specifically, this paper is a rejoinder to a Comment by Chelikowsky which is critical of our efforts to understand compound formation on the basis of self-consistent electronic-structure calculations. Our study, which was confined to a restricted class of compounds, concluded that Miedema's empirical rules work well, but that the physical model he proposes to justify the rules is not correct. Chelikowsky contends that the set of compounds on which we base this conclusion is small and unrepresentative and that our rationalization of Miedema's success using an incorrect model appears to fail, when applied to a broader class of compounds. We argue here that: (1) the class of compounds we consider is the entire set of compounds for which the connection between Miedema's rules and his physical model is unambiguous, (2) the physical picture of the compound-formation process that we advocate is more conventional and more fundamentally plausible than Miedema's, and (3) with regard to the technique we employ to rationalize Miedema's success with an incorrect model, Chelikowsky has simply misused the underlying concept. © 1982 The American Physical Society.