One hundred and sixty-three patients subjected to three different types of thoracic operation were allocated randomly either to balanced intravenous anaesthesia including i.v. opiates with post-operative intramuscular opiates (intramuscular group) or to balanced intravenous anaesthesia without i.v. opiates but with high thoracic epidural regional block during the operation as well as epidural nicomorphine post-operatively (epidural group). Post-operative nicomorphine was given only at the request of the patient and as frequently as needed to obtain satisfactory pain relief. Patients in the epidural group received nicomorphine exclusively by epidural injection. In this paper (Part I), general parameters between the epidural group and the im. group are compared. The epidural group showed acceptable cardiovascular stability during the operation. With the catheter tip at the T3-T4 level, there was apparently a preferential spread downwards of the local anaesthetics. Post-operatively both groups (i.m. and epidural) provided effective and rapid onset of analgesia. There were no major differences in post-operative pain assessment, either by the patient or by the team.