To render rational something so seemingly irrational as the Stalinist terror goes against all natural instincts. Gregory, however, clinically (and by applying rational-choice models) does indeed argue that Joseph Stalin, as other tyrants before him, rationally calculated both the utility and the limits of terror as a means of buttressing his and his regime's power and furthering its agenda. He also explores the rationality in the way Stalin and his inner circle rejiggered the organization of the state security organs, the kinds of people they relied on to run them, and the incentive structures they created to ensure that the praetorians would obey and not turn on them. To this he adds an analysis of the difficult but rational matrix of choices the agents of terror had to face, along with the careful definition of the terror's victims as it was designed to stay within the bounds of what the rest of the country would find plausible. The data, a good deal of it made available only recently, are used well, and the case made makes the incomprehensible less so. Still, there remains that point at which terror "by quota" loses any sense.