A persuasive body of evidence indicates that substantial protection against chemical carcinogenesis can be achieved by induction of enzymes concerned with the metabolism of carcinogens. There are two classes of anticarcinogenic enzyme inducers: (a) monofunctional inducers (e.g., phenolic antioxidants, isothiocyanates, coumarins, thiocarbamates, cinnamates, 1,2-dithiol-3-thiones) that elevate Phase II enzymes (such as glutathione S-transferases, NAD(P)H:quinone reductase, UDP-glucuronosyl-transferases) in various tissues without significantly raising the Phase I enzyme, aryl hydrocarbon hydroxylase (cytochrome P1-450); and (b) bifunctional inducers (e.g., polycyclic aromatic hydrocarbons, flavonoids, and azo dyes) that induce both Phase I and Phase II enzymes of xenobiotic metabolism. Induction of Phase II enzymes appears to be a sufficient condition for achieving chemoprotection, and since certain Phase I enzymes are responsible for activating carcinogens to their ultimate reactive forms, selective Phase II enzyme inducers offer intrinsically safer prospects for achieving chemoprotection. Whereas induction of both Phase I and II enzymes by bifunctional inducers depends on the Ah receptor, induction of Phase II enzymes by monofunctional inducers is independent of a functional Ah receptor. Studies on the structural requirements for induction of quinone reductase [NAD(P)H:(quinone acceptor) oxidoreductase; EC 1.6.99.2] by monofunctional inducers in Hepa 1c1c7 murine hepatoma cells have revealed that such inducers contain a distinctive chemical feature (or acquire this feature by metabolism) that regulates the synthesis of this protective enzyme. The inducers are all Michael reaction acceptors characterized by olefinic (or acetylenic) linkages that are rendered electrophilic by conjugation with electron-withdrawing groups. Typical examples are alpha, beta-unsaturated aldehydes, ketones (including quinones), thioketones, sulfones, esters, nitriles and nitro groups. The potency of these inducers parallels their reactivity as Michael acceptors. These generalizations have provided mechanistic insight into the vexing question of how so many seemingly unrelated anticarcinogens induce chemoprotective enzymes. They have also led to the prediction of entirely new and unsuspected structures of inducers, with potential for chemoprotective activity.