Cocaine's ability to interact with sigma receptors suggests that these proteins mediate some of its behavioral effects. Therefore, three novel sigma receptor ligands with antagonist activity were evaluated in Swiss Webster mice: BD1018 (3S-1-[2-(3,4-dichlorophenyl)ethyl]-1,4-diazabicyclo[4.3.0]nonane), BD1063 (1-[2-(3,4-dichlorophenyl)ethyl]-4-methylpiperazine), and LR132 (1R,2S-(+)-cis-N-[2-(3,4-dichlorophenyl)ethyl]-2-(1-pyrrolidinyl)cyclohexylamine). Competition binding assays demonstrated that all three compounds have high affinities for sigma1 receptors. The three compounds vary in their affinities for sigma2 receptors and exhibit negligible affinities for dopamine, opioid, GABA(A) and NMDA receptors. In behavioral studies, pre-treatment of mice with BD1018, BD1063, or LR132 significantly attenuated cocaine-induced convulsions and lethality. Moreover, post-treatment with LR132 prevented cocaine-induced lethality in a significant proportion of animals. In contrast to the protection provided by the putative antagonists, the well-characterized sigma receptor agonist di-o-tolylguanidine (DTG) and the novel sigma receptor agonist BD1031 (3R-1-[2-(3,4-dichlorophenyl)ethyl]-1,4-diazabicyclo[4.3.0]nonane) each worsened the behavioral toxicity of cocaine. At doses where alone, they produced no significant effects on locomotion, BD1018, BD1063 and LR132 significantly attenuated the locomotor stimulatory effects of cocaine. To further validate the hypothesis that the anti-cocaine effects of the novel ligands involved antagonism of sigma receptors, an antisense oligodeoxynucleotide against sigma1 receptors was also shown to significantly attenuate the convulsive and locomotor stimulatory effects of cocaine. Together, the data suggests that functional antagonism of sigma receptors is capable of attenuating a number of cocaine-induced behaviors.