Manuscript received 5 April 2002. Rydelek and Sacks (2003) (R&S) have criticized our mapping of the minimum magnitude of completeness, M c, in earthquake catalogs (Wiemer and Wyss, 2000). They suggest that their own approach, based on Schuster's (1897) method, is superior. Clearly, tracking down variations in M c as a function of three-dimensional space and time is a complex task, and every method of attempting to do this has advantages, paid for by shortcomings. We maintain our position that, in the balance of advantages and shortcomings, our method is preferable. While we agree with the concluding statement of RS Taylor et al. , 1990): consequently, our interpretation remains that the breakdown in earthquake scaling that R&S claimed to see a decade ago is an artifact, and one should not draw conclusions about earthquake physics from it. R&S's first objection is that, in some cases, the assumption of a linear FMD is not valid. While this is true under unusual conditions, we disagree with R&S's conclusion that our method is incapable of finding evidence for such deviations. In fact, in our article we devoted an entire section to this topic and identified several regions where such a breakdown occurs (figures 4b, d, and f of Wiemer and Wyss, 2000). These regions are generally identified as regions with a poor goodness of fit or a sudden jump in completeness. Once identified, these regions can be specifically investigated to find the reason for the deviation (e.g., quarry blasts, temporal variations in M …