Protease HslV and ATPase HslU form an ATP-dependent protease in bacteria. We have previously determined the structure of the components of this protease. In the case of HslU, the structure was derived from HslU-HslV cocrystals, combining phase information from MAD and the previously determined HslV model. Whereas the structures of the components were confirmed in detail by later structures, the quaternary arrangement of HslV and HslU was not reproduced in later crystal forms. In a recent communication to this journal, Wang attempted a reinterpretation of our original data to account for this difference. In response, we demonstrate that difference Pattersons, difference Fouriers, molecular replacement calculations, R factors, and omit maps all support our original analysis and prove that the suggested reinterpretation is false by these criteria. In particular, we show that our crystals are essentially untwinned and that only the originally reported quaternary arrangement of HslV and HslU particles is consistent with the experimental data. We finally demonstrate that Wang's newly introduced Rtpa,, method to predict translational corrections for a subset of the unit cell contents is systematically flawed.