Stimulants, illegal and legal, continue to be used in competitive sport. The evidence for the ergogenic properties of the most potent stimulants, amphetamines, cocaine and ephedrine, is mostly insubstantial. Low doses of amphetamines may aid performance where effects of fatigue adversely affect higher psychomotor activity. Pseudoephedrine, at high doses, has been suggested to improve high intensity and endurance exercise but phenylpropanolamine has not been proven to be ergogenic. Only caffeine has substantial experimental backing for being ergogenic in exercise. The mode of action of these stimulants centres on their ability to cause persistence of catecholamine neurotransmitters, with the exception of caffeine which is an adenosine receptor antagonist. By these actions, the stimulants are able to influence the activity of neuronal control pathways in the central (and peripheral) nervous system. Rodent models suggest that amphetamines and cocaine interact with different pathways to that affected by caffeine. Caffeine has a variety of pharmacological effects but its affinity for adenosine receptors is comparable with the levels expected to exist in the body after moderate caffeine intake, thus making adenosine receptor blockade the favoured mode of ergogenic action. However, alternative modes of action to account for the ergogenic properties of caffeine have been supported in the literature. Biochemical mechanisms that are consistent with more recent research findings, involving proteins such as DARPP-32 (dopamine and cAMP-regulated phosphoprotein), are helping to rationalize the molecular details of stimulant action in the central nervous system.