While in vitro results at clinically relevant concentrations do not predict abacavir (1592U89) interactions with drugs highly metabolized by cytochrome P450, the potential does exist for a pharmacokinetic interaction between abacavir and ethanol, as both are metabolized by alcohol dehydrogenase. Twenty-five subjects were enrolled in an open-label, randomized, three-way-crossover, phase I study of human immunodeficiency virus-infected male subjects. The three treatments were administration of (i) 600 mg of abacavir, (ii) 0.7 g of ethanol per kg of body weight, and (iii) 600 mg of abacavir and 0.7 g of ethanol per kg. Twenty-four subjects completed the study with no unexpected adverse events reported. Ethanol pharmacokinetic parameters were unchanged with abacavir coadministration. The geometric least squares mean area under the concentration curve extrapolated to infinite time for abacavir increased 41% (from 11.07 to 15.62 microg. h/ml), and the half-life increased 26% (from 1.42 to 1.79 h) in the presence of ethanol (mean ethanol maximum concentration in plasma of 498 microg/ml). The percentages of abacavir dose recovered in urine as abacavir and its two major metabolites were each altered in the presence of ethanol, but there was no change in the total percentage ( approximately 50%) of administered dose recovered in the 12-h collection interval. In conclusion, while a single 600-mg dose of abacavir does not alter blood ethanol concentration, ethanol does increase plasma abacavir concentrations.