Recent studies in normotensive Wistar-Kyoto (WKY) and spontaneously hypertensive (SHR) rats show that some nonlinear indexes derived from the recurrence plot method may be better markers of sympathetic activity than the spectral powers of blood pressure (BP). We herein investigated the relationships between nonlinear indexes and plasma noradrenaline concentration in conscious WKY rats and SHRs. Blood pressure was recorded for 30 min after intravenous injection of saline (0.9% NaCl, 100 microL/kg), hexamethonium (20 mg/kg), atropine (0.5 mg/kg), atenolol (1 mg/kg) or prazosin (1 mg/kg). Spectral power in the low-frequency (LF) band and the nonlinear index (L(max)), calculated on diastolic (DBP) and systolic blood pressures (SBP), were used to analyse the BP signal. Noradrenaline concentration was determined by radioenzymatic technique. A robust stepwise regression analysis - using noradrenaline concentration as dependent variable, and LF, L(max) and treatment, as independent variables -shows that treatment is the main variable explaining the variance of noradrenaline level in WKY rats, excluding the use of the pooled data to explore the relationship between noradrenaline concentration and LF or L(max). In contrast, in SHRs, treatment has no effect on the variance of noradrenaline concentration and the pooled data were then used. In this group, no correlation was observed between noradrenaline concentration and LF. In contrast, very high positive correlation was observed between noradrenaline level and L(max)-DBP (r = 0.59; P = 0.0005) or L(max)-SBP (r = 0.53; P < 0.002). The results strengthen our previous suggestion that nonlinear indexes may be better tools than spectral powers to investigate the sympathetic nervous system.