Epidural and spinal anaesthesia results from the interaction of local anaesthetics with nerve structures, primarily those located within the subarachnoid space. Local anaesthetics can reach the sites of action along various distribution pathways. Uptake into extraneural tissues (in particular epidural fat) and systemic absorption compete with neural tissue distribution and thereby affect the clinical potency and duration of action. Consequently, epidural doses must be much higher than spinal doses. The systemic absorption of lignocaine (lidocaine), bupivacaine and etidocaine following lumbar epidural administration has been shown to be biphasic, with a rapid initial absorption phase followed by a much slower absorption phase. Initial absorption rates of lignocaine and bupivacaine following subarachnoid injection are much slower, but the late absorption rates are similar to those after epidural administration. The tissue distribution characteristics of various amide-type agents are similar, because more extensive plasma binding offsets the greater tissue affinity of the more lipophilic compounds bupivacaine and etidocaine. The amide-type agents are predominantly eliminated by hepatic metabolism, except prilocaine, which is also metabolised elsewhere in the body. Ester-type agents are rapidly hydrolysed in blood and liver and are eliminated much faster than amide-type agents. The blood concentrations attained depend primarily upon the dose administered. The addition of adrenaline (epinephrine) reduces the peak plasma drug concentrations; similarly, the age of the patient, disease states and drug interactions may alter the pharmacokinetics to various extents. Because of the low dose requirements, systemic toxicity is not a problem during spinal anaesthesia. During epidural anaesthesia, however, the safety margin is relatively small, and systemic toxicity is very likely to occur after inadvertent intravascular injection of an epidural dose.