The purpose of this study was to identify the relationship between sensitivity of arterial baroreflex and plasma concentrations of lidocaine. Using twelve mongrel dogs anesthetized with alpha-chloralose, the left kidney was exposed retroperitoneally, and renal sympathetic nerve activity was recorded continuously. Lidocaine was infused in four different doses: 2 mg.kg BW-1 bolus + 100 micrograms.BW-1 x min; 3 mg.kg BW-1 bolus + 200 micrograms.kg BW-1 x min; 6 mg.kg BW-1 bolus + 400 micrograms.kg BW-1 x min; and 12 mg.kg BW-1 + 800 micrograms.kg BW-1 x min. Baroreflex depressor and pressor tests using sodium nitroprusside (5-10 micrograms.kg-1) and phenylephrine (2-4 micrograms.kg-1) were performed before and at 10 min after beginning lidocaine infusion. Plasma lidocaine concentrations determined by high performance liquid chromatography revealed that the steady-state levels were maintained during the baroreflex tests. Baroreflex sensitivity was preserved at plasma concentrations of lidocaine below 5 micrograms.ml-1. However, cardiac and sympathetic baroreflex sensitivity were significantly attenuated (P < 0.01) when plasma lidocaine concentrations were well above human convulsion levels (10 micrograms.ml-1). The results indicate that hemodynamic derangement observed in the lidocaine-induced central nervous system toxicity is, at least in part, due to the attenuated arterial baroreflex.