Serotonin transporter has a key-role in regulation of serotoninergic function, and is involved in numerous neurodegenerative and psychiatric disorders. To obtain an efficient radioactive ligand allowing the study of this transporter in vitro and in vivo, we synthesized a new diphenyl sulfide derivative, N,N-dimethyl-2-(2-amino-4-methylphenylthio)benzylamine or MADAM. We present here extensive pharmacological characterization of this compound. [3H]MADAM bound to serotonin transporters with a very high affinity in vitro on rat cortical membranes, at least 2 times better than the most commonly used radioactive probes (Kd, 60 pM; Bmax, 543 fmol/mg of protein). Competition studies showed few inhibitory effect of nisoxetine (Ki = 270 nM), no inhibitory effect of desipramine or 1-[2-(diphenylmethoxy) ethyl]-4-(3-phenylpropyl)piperazine (GBR 12935) (Ki >1000 nM), and strong effect of paroxetine (Ki = 0.32 nM) and citalopram (Ki = 1.57 nM). Therefore, MADAM has around 1000-fold better selectivity for the serotonin transporter than for other transporters. Autoradiographic studies both on rat and postmortem human brain slices demonstrated that the distribution of [3H]MADAM parallels the localization of serotonin transporters and is prevented by known inhibitors of them. The high affinity and selectivity of [3H]MADAM for the serotonin transporter show that it is very valuable for studies using in vitro approaches. The high selectivity and low nonspecific binding of [3H]MADAM on the postmortem human brain, together with preliminary in vivo results with [11C]MADAM, is a new argument for future use of this ligand in in vivo studies of the distribution, pharmacology, and pathophysiology of the serotonin transporter in the human brain with positron emission tomography.