Central stimulants readily serve as training stimuli in drug discrimination studies and typically substitute for one another in tests of stimulus generalization regardless of which is used as training drug. We have previously found that, although substitution occurs between (+)amphetamine and (-)ephedrine, substitution did not occur upon administration of S(+)methamphetamine to (-)ephedrine-trained animals. In the present investigation, rats were trained to discriminate S(+)methamphetamine (1 mg/kg) from saline vehicle and tests of stimulus generalization were performed with several stimulants, including (-)ephedrine. The S(+)methamphetamine stimulus (ED(50)=0.06 mg/kg) generalized to R(-)methamphetamine (ED(50)=1.61 mg/kg), S(+)amphetamine (ED(50)=0.28 mg/kg), S(-)methcathinone (ED(50)=0.21 mg/kg), methylphenidate (ED(50)=0.28 mg/kg), cocaine (ED(50)=3.68 mg/kg) and (-)ephedrine (ED(50)=13.1 mg/kg). Hence, stimulus generalization between S(+)methamphetamine and (-)ephedrine is apparently asymmetrical. In a companion study, R(-)methamphetamine was administered to rats trained to discriminate (-)ephedrine (4 mg/kg); substitution occurred and R(-)methamphetamine (ED(50)=0.92 mg/kg) was found to be nearly equipotent with (-)ephedrine (ED(50)=0.8 mg/kg). Although the exact basis for the observed results are unclear, they are discussed in terms of the different effects of (-)ephedrine and the methamphetamine optical isomers on neurotransmitter release and reuptake.