The current study examined the possibility that high local anesthetic concentrations within the central nervous system could contribute to the toxic cardiovascular effects observed clinically after an accidental intravenous injection. Equal numbers of molecules of lidocaine (1.6 microliter of a 2% solution) and bupivacaine (1.0 microliter of a 4% solution) were injected at three vasomotor and cardioactive areas in the rat medulla by means of a 28-gauge hypodermic needle and a microsyringe. These areas were the C1 region, the intermediolateral column (IML), and the nucleus tractus solitarius (NTS). Both lidocaine and bupivacaine at C1 significantly decreased mean arterial pressure and, at IML, resulted in significant bradycardia and hypotension. At NTS, both lidocaine and bupivacaine caused significant bradycardia and hypotension, which were accompanied by ventricular arrhythmias in 55% of the animals. In all animals in whom ventricular arrhythmias were associated with lidocaine, the arrhythmias spontaneously reverted to normal sinus rhythm. In 50% of animals developing ventricular arrhythmias after bupivacaine, the arrhythmias resulted in death. Using an equal number of molecules of lidocaine and bupivacaine, all three regions studied demonstrated that bupivacaine was 2-4 times more potent than lidocaine in producing cardiovascular effects. These data demonstrate that direct application of local anesthetics within the medullary region of the central nervous system can result in hypotension, bradycardia, and ventricular arrhythmias similar to what may be seen in humans after accidental intravenous injections of local anesthetics. Moreover, the sites and mechanisms of action appear to be identical for bupivacaine and lidocaine. Thus bupivacaine does not appear to be an aberrant local anesthetic, as some have suggested, but rather produces more profound effects related to its potency and physicochemical properties.