Despite the success of highly active antiretroviral therapy (HAART) regimens in reducing HIV morbidity and mortality on the population level, individual responses and adverse effects are variable. Many factors influence the outcome of HAART, including the amount of drug that reaches the site of activity (treatment exposure). Behavioral factors and pharmacological parameters (bioavailability, clearance, distribution) influence exposure. Since antiretroviral drugs must be present in adequate quantities for long periods to be effective, measures of long-term exposure are needed. Stimulated by the finding that indinavir (an HIV protease inhibitor) levels in hair were highly correlated with viral load responses and resistance, we initiated preliminary work to develop methods for quantifying the amount of antiretroviral protease inhibitors and non-nucleoside reverse transcriptase inhibitors in small samples of hair. We found a strong correlation between lopinavir levels in hair and initial virologic response to lopinavir-based therapy; lopinavir concentrations in hair were significantly higher in responders versus nonresponders (Mann-Whitney exact test, p < 0.0001). Hair levels of antiretrovirals may correlate more closely with long-term outcomes than single blood levels. We use another chronic illness, diabetes mellitus, to compare the value of short term and longer-term measures of disease control (blood glucose and hemoglobin Ale, respectively) as an analogy for the comparison of antiretroviral levels in blood and in hair. The overarching goals of the proposed research are to develop and evaluate methods for determining long-term exposure to antiretroviral drugs in hair samples and to test the predictive value of the hair measures for virologic and immunologic responses, as well as the incidence of adverse effects and resistance mutations. We also propose to assess the value of correcting drug levels in hair by eumelanin content (a measure of the binding capacity for drug). Measuring drug levels in hair has been successfully applied to determination of exposure to a broad array of therapeutic and recreational drugs. This work may provide an inexpensive, noninvasive, field-friendly method for determination of the level of antiretroviral exposure, which could be applied in domestic and international AIDS treatment and research settings. We therefore propose a series of studies of the value of antiretroviral drug measurement using hair specimens based in the Women's Interagency HIV Study (WIHS) cohort, the largest longitudinal cohort study of HIV infection in women in the U.S.