Regional anesthesia has become a routine part of the practice of anesthesiology in infants and children. Local anesthetic toxicity is extremely rare in infants and children; however, seizures, dysrhythmias, cardiovascular collapse, and transient neuropathic symptoms have been reported. Infants and children may be at increased risk from local anesthetics compared with adults. Larger volumes of local anesthetics are used for epidural anesthesia in infants and children than in adults. Metabolism and elimination of local anesthetics can be delayed in neonates, who also have decreased plasma concentrations of alpha(1)-acid glycoprotein, leading to increased concentrations of unbound bupivacaine. Most regional anesthetic procedures in infants and children are performed with the patient heavily sedated or anesthetized; because of this, and because a test dose is not a particularly sensitive marker of intravenous injection in the anesthetized patient, detection of intravascular local anesthetic injection is extremely difficult. The same local anesthetics used in adult anesthetic practice are also used in infants and children. Because of its extremely short duration of action, chloroprocaine has been used primarily for continuous epidural techniques in infants and children. The use of tetracaine has generally been limited to spinal and topical anesthesia. Lidocaine (lignocaine) has been used extensively in infants and children for topical, regional, plexus, epidural and spinal anesthesia. The association between prilocaine and methemoglobinemia has generally restricted prilocaine use in infants and children to the eutectic mixture of local anesthetics (EMLA). Because of its greater degree of motor block compared with other long-acting local anesthetics, etidocaine has generally been limited to plexus blocks in infants and children. Mepivacaine has been used for both plexus and epidural anesthesia in infants and children. Because postoperative analgesia is often the primary justification for regional anesthesia in infants and children, bupivacaine, a long-acting local anesthetic, is the most commonly reported local anesthetic for pediatric regional anesthesia. Given the lower toxic threshold of bupivacaine compared with other local anesthetics, the risk-benefit ratio of bupivacaine may be greater than that of other local anesthetics. Two new enantiomerically pure local anesthetics, ropivacaine and levobupivacaine, offer clinical profiles comparable to that of bupivacaine but without its lower toxic threshold. The extreme rarity of major toxicity from local anesthetics suggests that widespread replacement of bupivacaine with ropivacaine or levobupivacaine is probably not necessary. However, there are clinical situations, including prolonged local anesthetic infusions, use in neonates, impaired hepatic metabolic function, and anesthetic techniques requiring a large mass of local anesthetic, where replacement of bupivacaine with ropivacaine, levobupivacaine or (for continuous techniques) chloroprocaine appears prudent.