Background: Quercetin is a naturally occurring flavonoid with many biological activities including inhibition of a number of tyrosine kinases. A phase I, dose-escalation trial of quercetin defined the maximum tolerated dose (MTD) as 1700 mg/m(2) three weekly, but the vehicle, dimethyl sulphoxide (DMSO) is unsuitable for further clinical development of quercetin. Patients and methods: A water-soluble, pro-drug of quercetin (3'(N-carboxymethyl)carbomyl-3,4',5,7-tetrahydroxyflavone), QC12 has been synthesised. Six cancer patients received 400 mg of QC12 (equivalent to 298 mg of quercetin), orally on day 1 and intravenously (i.v.) in normal saline on day 14. Results: Following oral administration of QC12 we were unable to detect QC12 or quercetin in plasma. After i.v. administration, we detected peak plasma concentrations of QC12 of 108.7 +/- 41.67 mu Molar (muM). A two-compartment model with mean t(1)/2 alpha of 0.31 +/- 0.27 hours and mean t(1)/2 beta of 0.86 +/- 0.78 hours best described the concentration-time curves for QC12. The mean AUC was 44.54 +/- 13.0 muM.hour and mean volume of distribution (Vd) of 10.0 +/- 6.2 litres (l). Quercetin was found in all patients following i.v. infusion of QC12, with peak levels of quercetin 19.9 +/- 11.8 muM. The relative bioavailability of quercetin was estimated to be 20%-25% quercetin released from QC12. Conclusions: QC12 is not orally bioavailable. This water-soluble pro-drug warrants further clinical investigation; starting with a formal phase I, IV, dose-escalation study.