Cocaine's blockade of dopamine reuptake by brain dopamine transporters (DAT) is a central feature of current understanding of cocaine reward and addiction. Empirical screening of small-molecule chemical libraries has thus far failed to provide successful cocaine blockers that allow dopamine reuptake in the presence of cocaine and provide cocaine "antagonism". We have approached this problem by assessing expression, dopamine uptake, and cocaine analog affinities of 56 DAT mutants in residues located in or near transmembrane domains likely to play significant roles in cocaine recognition and dopamine uptake. A phenylalanine-to-alanine mutant in putative DAT transmembrane domain 3, F154A, retains normal dopamine uptake, lowers cocaine affinity 10-fold, and reduces cocaine stereospecificity. Such mutants provide windows into DAT structures that could serve as targets for selective cocaine blockers and document how combined strategies of mutagenesis and small molecule screening may improve our abilities to identify and design compounds with such selective properties.