1 To determine the site of action of rilmenidine, we examined its effects on arterial blood pressure (BP), heart rate (HR) and postganglionic renal sympathetic nerve activity (RSNA) after intracerebroventricular (i.c.v.) administration (300pugkg-1), in groups (all n = 6) of conscious and freely moving, pentobarbitoneanaesthetized and pentobarbitone-anaesthetized and spinally transected, fifteen week-old male spontaneously hypertensive rats (SHRs). 2 In conscious SHRs, which exhibited a low sympathetic nerve activity (RSNA: 3.4 + 0.9,uV), rilmenidine was inactive on systolic BP (SBP), diastolic BP (DBP), HR and RSNA. 3 In intact pentobarbitone-anaesthetized SHRs, which exhibited an elevated sympathetic nerve activity (RSNA: 10.6 + 0.9 pV), rilmenidine exerted potent antihypertensive (ASBP: -37 + 4%; ADBP:-43 + 6%), bradycardic (AHR: -32 + 3%) and sympathoinhibitory (ARSNA: -68 + 9%) activities. cut-off set at 100 and 1,000 Hz, respectively. The amplified nerve activity was displayed on a Gould oscilloscope DSO-1602 (Gould Instr. S.A.F., Paris, France). For the integration of discharges the nerve signals appearing at the oscilloscope outputs were rectified and displayed along with arterial BP and HR on a computer MOTOROLA SYS-3304/NY/082 (Motorola Co., Tempe, U.S.A.) (Gustin et al., 1990). The background non neural activity was taken as the activity recorded after injection of a ganglionic blocker (trimethaphan camsilate: 10mgkg-1, i.v.). This background activity was determined at the end of the experiment and substracted from the total integrated activity to allow the quantitative estimation of sympathetic nerve activity. and spike activity was monitored. After a 2 h period which allowed the animal to recover from anaesthesia and the cardiovascular parameters to stabilize, BP, HR and RSNA were continuously recorded for 5 to 6 h. After a 1 h control period, each animal received one i.c.v. injection of artificial cerebrospinal fluid (CSF) followed by one i.c.v. injection of rilmenidine (300 jig kg 1).Pentobarbitone-anaesthetized SHRs (n = 6) After connection to the BP recording system, implantation of the renal nerve bipolar electrode and cannulation of the trachea, each animal was placed in a stereotaxic frame, for an equilibration period which allowed the cardiovascular parameters, blood gases and pH to stabilize. Then, BP, HR and RSNA were recorded continuously for a 1 h control period and a 3-4 h period during which, each animal received one i.c.v. injection of artificial CSF followed by one of rilmenidine (300 ,ug kg-1) Pentobarbitone-anaesthetized and spinally-transected SHRs (n = 6) The animals were prepared as described above for anaesthetized SHRs. Then, they were mounted on a stereotaxic frame in order to hold the head steady whilst the spinal cord was dissected and BP, HR and RSNA were continuously recorded. When the cardiovascular parameters returned to control values, the spinal cord was completely transected at C5 level and they were monitored for a 3-4h period. Twenty to thirty min after transection,...