1. The pharmacokinetics of hydroxychloroquine were studied in five healthy volunteers following an intravenous infusion of 155 mg (2.47 +/- 0.25 mg kg-1) racemic hydroxychloroquine. Four of these volunteers also received a further 310 mg (4.92 +/- 0.45 mg kg-1) infusion of hydroxychloroquine and evidence of nonlinearities in the pharmacokinetics of hydroxychloroquine were sought. 2. No nonlinear elimination or distribution processes appeared to be operating at the doses of hydroxychloroquine used in this study, supporting the hypothesis that in the therapeutic dosing range the pharmacokinetics of hydroxychloroquine are linear. 3. Half-life and mean residence time were long (around 40 days) and large volumes of distribution were calculated (5,522 l from blood, 44,257 l from plasma). Sequestration into the tissues is an important feature of the disposition of hydroxychloroquine. The persistence of hydroxychloroquine in the body is due primarily to this extensive tissue distribution, rather than to low clearance (667 ml min-1 based on plasma data, 96 ml min-1 based on blood data). 4. Plasma data were more variable than blood data. Blood to plasma concentration ratios were not constant (mean +/- s.d.: 7.2 +/- 4.2). The data indicate that it is preferable to measure whole blood concentrations of hydroxychloroquine, rather than plasma concentrations, in pharmacokinetic studies. 5. The pharmacokinetics of hydroxychloroquine are similar to those of chloroquine.