The influence of melanin on the binding of xenobiotics in hair will impact the interpretation of drug concentrations determined by hair testing. The purpose of this study was to determine if codeine, as a model compound of abused drugs, would be incorporated into black, brown, blond, or red hair as a function of melanin concentration. Such data would assist in the interpretation of codeine concentrations in hair and help elucidate the potential influence of hair color on incorporation of drugs. Male and female Caucasians with black (n = 6), brown (n = 12), blond (n = 8), or red hair (n = 6) and non-Caucasians with black hair (n = 12) aged 21-40 years were enrolled in the study. Each subject was administered oral codeine phosphate syrup in a dosage of 30 mg three times a day for five days. Twenty-four hours after the end of the treatment period, a 30-mg codeine dose was administered and the subject's plasma area under the concentration time curve (AUC) for codeine was determined. Codeine and melanin were measured in the first 3 cm of hair closest to the vertex region of the scalp prior to and 1, 4, 5, 6, and 7 weeks after dosing. The quantitative and qualitative melanin profiles were determined for each subject's hair to provide an objective measure of hair color. The plasma concentrations of codeine were measured to eliminate differences in the bioavailability and clearance of codeine as factors that might account for the differences in codeine hair concentrations. The subjects were asked not to cut their hair in the vertex region of the scalp or to use any form of chemical treatment on their hair, but otherwise normal hygienic measures were permitted. The mean (+/- SE) hair codeine concentrations 5 weeks after dosing were 1429 (+/- 249) pg/mg in black hair; 208 (+/- 17) pg/mg in brown hair; 99 (+/- 10) pg/mg in blond hair; and 69 (+/- 11) in red hair pg/mg. In black hair, codeine concentrations were 2564 (+/- 170) pg/mg for Asians and 865 (+/- 162) pg/mg for Caucasians. Similar concentration relationships were observed at weeks 4, 6, and 7. A strong relationship between the hair concentrations of codeine and melanin (R(2) = 0.73) was observed. Normalization of the codeine concentration with the melanin concentration reduced the hair color differences observed. These data demonstrate that the interpretation and reporting of hair test results for codeine are influenced by hair color. After this dosing protocol, the proposed federal guideline cutoff of 200 pg/mg of codeine would result in 100% of subjects with black hair and 50% of subjects with brown hair being reported as positive, and subjects with blond or red hair would be reported as negative. The incorporation of these drugs into hair should be studied carefully in humans to ensure the appropriate interpretation of drug concentrations.