Chronic administration of lipophilic drugs can result in accumulation and prolonged elimination during abstinence. It has been suggested that cocaine and/or metabolites can be detected in saliva and urine for an extended period following long-term, high-dose administration. The effects of chronic oral cocaine administration in healthy volunteer subjects with a history of cocaine abuse were investigated. Subjects were housed on a closed clinical ward and were administered oral cocaine in up to 16 daily sessions. In each session, volunteers received five equal doses of oral cocaine with 1 h between doses. Across sessions, cocaine was administered in ascending doses from an initial dose of 100 mg (500 mg/day) up to 400 mg (2 g/day), increasing by 25 mg/dose/session (125 mg/session). Participation in the study was terminated if cardiovascular safety parameters were exceeded. Plasma and saliva specimens were collected periodically during the dosing sessions and during the one-week withdrawal phase at the end of the study. All urine specimens were collected throughout the entire study. Specimens were analyzed for cocaine and metabolites by solid-phase extraction followed by gas chromatographic-mass spectrometric analysis in the SIM mode. The limit of detection for each analyte was approximately 1 ng/mL. The analytes measured included benzoylecgonine (BZE), ecgonine methyl ester, cocaine, benzoylnorecgonine, norcocaine, m- and p-hydroxycocaine, and m- and p-hydroxybenzoylecgonine. Noncompartmental analysis was employed for the determination of plasma and saliva pharmacokinetic parameters. Urinary elimination half-lives for cocaine and metabolites were determined by constructing ARE (amount remaining to be excreted) plots. Two phases of urinary elimination of cocaine and metabolites were observed. An initial elimination phase was observed during withdrawal that was similar to the elimination pattern observed after acute dosing. The mean (N = 6) plasma, saliva, and urine cocaine elimination half-lives were 1.5 +/- 0.1 h, 1.2 +/- 0.2 h, and 4.1 +/- 0.9 h, respectively. For three subjects, the mean cocaine urinary elimination half-life for the terminal phase was 19.0 +/- 4.2 h. There was some difficulty in determining if a terminal elimination phase for cocaine was present for the remaining three subjects because of interference by high concentrations of BZE. A terminal elimination phase was also observed for cocaine metabolites with half-life estimates ranging from 14.6 to 52.4 h. These terminal elimination half-lives greatly exceeded previous estimates from studies of acute cocaine administration. These data suggest that cocaine accumulates in the body with chronic use resulting in a prolonged terminal elimination phase for cocaine and metabolites.