A simple technique for cannulation of the thoracic epidural space in rats was described. 40-50 microliter of epidural bupivacaine 5 mg/ml induced a distribution of sensory analgesia from lower cervical to lower thoracic segments. With this model, effects of thoracic epidural anaesthesia (TEA) on mean arterial pressure (MAP), cardiac output (CO), systemic vascular resistance (SVR), stroke volume (SV), heart rate (HR), central venous pressure (CVP), left ventricular end-diastolic pressure (LVEDP) and maximal increase of pressure in the left ventricle (max dp/dt) were studied in six groups of animals: 1) In conscious animals (n = 10) MAP, CO, SV and HR decreased significantly by 12%, 25%, 10% and 16%, respectively, while SVR increased significantly by 20% during TEA; 2) In chloralose-anaesthetized animals (n = 7) the reduction in CO during TEA was less pronounced and there were no significant changes in SV or SVR; 3) In conscious animals (n = 6) LVEDP, CVP and max dp/dt decreased significantly during TEA; 4) Hexamethonium, when administered to pharmacologically vagotomized conscious animals during TEA (n = 8), induced a significant decrease in SVR (23%) but no change in HR; 5) Changes in haemodynamics after cardiac adrenoceptor blockade with metoprolol, in conscious animals (n = 12), did not differ significantly from those seen during TEA, except for an unchanged SV after metoprolol; 6) 50 microliters of bupivacaine (5 mg/ml) when given i.v. to conscious animals (n = 8) did not affect CO, SV, HR or TPR significantly, while MAP increased slightly but significantly. Thus, in this conscious animal model, TEA almost completely and rather selectively blocked probably mediated by a reflex activation of unblocked sympathetic efferents.(ABSTRACT TRUNCATED AT 250 WORDS)