The analysis of hair has been proposed as a tool for monitoring drug-treatment compliance. This study was performed to determine if buprenorphine (BPR) and norbuprenorphine (NBPR) could be detected in human hair after controlled administration of drug and to determine if segmental analysis of hair was an accurate record of the dosing history. Subjects with dark hair (six males, six females) received 8 mg sublingual BPR for a maximum of 180 days. Single hair collections were made once after BPR treatment and stored at -20 degrees C until analysis. Hair was aligned scalp-end to tip and then segmented in 3-cm sections. For this study, it was assumed that the mean hair growth rate was 1.0 cm/month. Deuterated internal standard was added to hair segments (2-20 mg of hair) and digested overnight at room temperature with 1 N NaOH. Specimens were extracted with a liquid-liquid procedure and analyzed by liquid chromatography-tandem mass spectrometry. The limits of quantitation for BPR and NBPR were 3 pg/mg and 5 pg/mg, respectively, for 20 mg of hair. BPR and NBPR concentrations were highest for all subjects in hair segments estimated to correspond to the subject's period of drug treatment. With one exception, NBPR was present in higher concentrations in hair than was the parent compound. BPR concentrations in hair segments ranged from 3.1 pg/mg to 123.8 pg/mg. NBPR concentrations ranged from 4.8 pg/mg to 1517.8 pg/mg. In one subject, BPR and NBPR were not detected in any hair segment. In some subjects, BPR and NBPR were detected in hair segments that did not correspond to the period of drug treatment, suggesting that drug movement may have occurred by diffusion in sweat and other mechanisms. The data from this study also indicate that there is a high degree of intersubject variability in measured concentration of BPR and NBPR in hair segments, even when subjects receive the same dose for an equivalent number of treatment days. Future prospective studies involving controlled drug administration will be necessary to evaluate whether hair can serve as an accurate historical record of variations in the pattern of drug use.