The ability to detect cocaine use/exposure by either hair or sweat analysis was compared in a random population of adults at a major US university. Sweat was obtained by wiping the forehead with a cosmetic puff containing isopropanol. Using cut-off levels for sweat of 2.2 ng cocaine/wipe and of hair of 0.05 ng cocaine/mg hair, sweat detected two times more cocaine use/exposure than did hair. Sweat analysis detected a use rate of 12% compared to a 6% rate by hair analysis, both greater than the 2% that would be expected in this population. The high rate of detection was surprising and suggests that use of, if not exposure to, cocaine is underreported. Controlled experiments showed that cocaine could remain on the skin for about 3 days after external exposure. At the current state of knowledge, sweat appears to measure both use and exposure. Nevertheless, sweat testing could be used in several scenarios (such as roadside driving while intoxicated) where the case of collection and testing of sweat could outweigh the passive exposure considerations. Cocaine concentrations in skin swabs > 15 ng/swab would appear to indicate recent use/exposure.