Abstract: In recent articles, I have advanced a number of criticisms of the market definition/redefinition paradigm, chief among them that market definition is impossible and counterproductive. First, there is no valid way to infer market power from market shares in redefined (non-homogeneous-goods) markets. Second, one cannot choose which market definition is superior without already having in hand one’s best estimate of market power, rendering the exercise pointless. Worse, market power inferences in the chosen market are inferior to the best estimate with which one began. After elaborating these points, this Essay applies them to the three main settings in which the hypothetical monopolist test is employed in various jurisdictions’ merger guidelines, showing this test to be counterproductive in every instance. Finally, it addresses reasons that some are nevertheless reluctant to abandon market definition altogether.