The recognition that long-acting local anaesthetics, particularly bupivacaine the de facto standard long-acting local anaesthetic, were disproportionately more cardiotoxic than their shorter-acting counterparts stimulated the development of the bupivacaine congeners, ropivacaine and levobupivacaine. These agents, like all local anaesthetics, can produce cardiotoxic sequelae by direct and indirect mechanisms that derive from their mode of local anaesthetic actions, i.e. inhibition of voltage-gated ion channels. While all local anaesthetics can cause direct negative inotropic effects, ropivacaine and levobupivacaine are less cardiotoxic than bupivacaine judging by the larger doses tolerated in laboratory animal preparations before the onset of serious cardiotoxicity (particularly electro-mechanical dissociation or malignant ventricular arrhythmias). Additionally, they are less toxic to the CNS than bupivacaine judging by the larger doses tolerated before the onset of seizures. This may be clinically important because CNS effects may be involved in the production of serious cardiotoxicity. Preclinical studies in humans are a 'blunt instrument' in their ability to distinguish significant differences between these drugs because of the relatively small doses that can be used. Nevertheless, available evidence from human studies corroborates the preclinical laboratory animal studies. Because clinically significant differences between these drugs are more quantitative than qualitative, i.e. toleration of a larger dose before manifestation of toxicity, we have concluded that these newer agents have a lower risk of causing serious cardiotoxicity than bupivacaine. Thus, compared with bupivacaine, the newer agents may be seen as 'safer', but they must not be regarded as 'safe'.