The serotonin transporter (SERT) is a target for many clinically significant drugs, such as cocaine, amphetamine, and antidepressants. The relationship between the structure of SERT and the binding of substrates and antagonists is virtually unknown, despite a large body of data describing the structure-activity relationships of transporter ligands. The cloning of multiple species homologs of SERT affords a unique opportunity for molecular comparisons to identify potential domains and residues involved in ligand recognition. We have conducted pharmacological comparisons of the cloned rat and human SERTs in transiently transfected HeLa cells. Serotonin uptake and radioligand binding assays revealed that rat and human SERTs show different sensitivities to some but not all transporter ligands; most tricyclic antidepressants were significantly more potent at the human SERT, relative to rat SERT, whereas d-amphetamine was a more potent inhibitor of rat SERT. Several other ligand such as fluoxetine, paroxetine, (+)-methylenedioxymethamphetamine, cocaine, and the substrate 5-hydroxytryptamine, shows no significant species selectivity. Cross-species chimeras between rat and human SERTs were constructed to track the species-specific pharmacologies through the SERT molecule. These chimeric SERTs were expressed in HeLa cells and transported serotonin similarly to parental SERTs. Using these chimeras, we have isolated a region distal to amino acid 532 the imparts species preferences for both the tricyclic imipramine and d-amphetamine. Our results support the prediction of distinct binding sites for SERT ligands and implicate a restricted region in or near putative transmembrane domain 12 of the transport as being involved in both substrate and antagonist recognition.