Sulforaphane (SF) is considered to be the major anticarcinogenic component in broccoli. The effects of feeding rats purified SF (5 mmol/kg of diet), broccoli containing SF formed in situ during laboratory hydrolysis (broccoli-HP; 20% freeze-dried broccoli diet, 0.16 mmol of SF/kg of diet), and broccoli containing intact glucosinolates (broccoli-GS; 20% freeze-dried broccoli diet, 2.2 mmol of glucoraphanin/kg of diet) were compared. Rats (male F344 rats, five per group) were fed control (modified AIN-76 B-40), SF, broccoli-HP, or broccoli-GS for 5 days. In rats fed broccoli-GS, quinone reductase activities (QR) in the colon and liver were greater (4.5- and 1.4-fold over control, respectively) than in rats fed broccoli-HP (3.2- and 1.1-fold over control, respectively). Broccoli-GS and SF diets increased QR to the same extent, even though the broccoli-GS diet contained far less SF (as the unhydrolyzed glucosinolate, glucoraphanin) than the purified SF diet. In a second experiment, rats were fed one of six diets for 5 days: (1) control; (2) 20% broccoli-GS; (3) diet 2 + low SF (0.16 mmol/kg of diet); (4) diet 2 + high SF (5 mmol/kg of diet); (5) low SF (0.32 mmol/kg of diet); or (6) high SF (5.16 mmol/kg of diet). In both liver and colon, QR was increased most by high SF plus broccoli-GS; individually, high SF and broccoli-GS had similar effects, and adding the low-dose SF to broccoli-GS had either no effect or a negative effect. In both experiments, urinary SF-mercapturic acid correlated with QR activity, not with dietary intake. It was concluded that all diets were substantially more effective in the colon than in the liver and that broccoli-GS was more potent than SF or broccoli-HP.