The lipophilic character of quercetin suggests that it can cross enterocyte membranes via simple diffusion. Therefore, it should be more bioavailable than its glucosides, which require preliminary hydrolysis or active transport for absorption. However, the published human studies show that quercetin is less bioavailable than its glucosides. Assuming that low bioavailability of quercetin aglycone provided to humans as a pure substance is the result of its low solubility in the digestive tract, we studied its bioavailability from dietary sources in which quercetin was dispersed in the food matrix. In a randomized crossover study, 9 volunteers took a single dose of either shallot flesh (99.2% quercetin glucosides and 0.8% quercetin aglycone) or dry shallot skin (83.3% quercetin aglycone and 16.7% quercetin glucosides), providing 1.4 mg quercetin per kg of body weight. Blood samples were collected before and after consumption of shallot preparations. Plasma quercetin was measured on HPLC with electrochemical detection after plasma enzymatic treatment. The maximum plasma quercetin concentration of 1.02 +/- 0.13 micromol/L was reached at 2.33 +/- 0.50 h after shallot flesh consumption compared with 3.95 +/- 0.62 micromol/L at 2.78 +/- 0.15 h after dry skin consumption. The area under the concentration-time curve after dry skin consumption was 47.23 +/- 7.53 micromol x h(-1) x L(-1) and was significantly higher than that after shallot flesh intake (22.23 +/- 2.32 micromol x h(-1) x L(-1)). When provided along with dietary sources, quercetin aglycone is more bioavailable than its glucosides in humans. Results point to the food matrix as a key factor.