We are reporting on postoperative pain treatment using epidural analgesia in 1,822 patients, performed between 1995 to 2000, following continuous epidural anaesthesia combined with general anaesthesia for operations in various specialized areas (general or visceral surgery, vascular and thoraxic surgery, gynaecology, urology and orthopaedics). A total of 1,727 of these postoperative epidurals were included in a detailed evaluation. The postoperative epidural analgesia consisted of a continuous application of 0.25% bupivacain or 0.2% ropivacain. These local anaesthetics were administered epidurally in an hourly perfusion rate of 7.5 ml. We found "good" pain relief through continuous epidural administering of the local anaesthetics in 1,292 patients (74.8%). "Moderate" pain relief was achieved in 392 patients (22.7%). Sufentanil had to be epidurally administered in addition to local anaesthetics in 262 patients (15.2%) in the wake-up room. The sufentanil doses lay between 5 and a maximum 10 micrograms per hour. An additional epidural application of morphine-boli in a dose of 3 mg every 8-12 hours was necessary in 384 patients (22.2%) in the surgical wake-up stations. In 392 patients (22.7%), the additional systemic administering of antipyretic analgesics such as metamizol or paracetamol or spasmolytica was sufficient. In 43 cases (2.5%), sufficient pain relief could not be achieved with epidural analgesia even with additive applications of systemic functioning pharmaceuticals, so that the postoperative pain therapy had to be completely switched to a PCA. The lying time of the epidural catheter was 2-5 days. It was shortest with the gynaecological patients and longest with patients from general, visceral, thoraxic and vascular surgery areas. An important factor for a sufficient epidural analgesia is the exact epidural positioning of the catheter tip in the area of the spinal cord segments, which are affected by the operation. This reveals the required puncture height. The following side-effects resulting from the epidural analgesia were found: blood pressure loss of more than 20% of the starting value (21%), temporary bladder voiding disorders (8%), temporary sensory disorders of the lower extremities (6.5%), seldom nausea (2.4%) and post-puncture headaches (1.2%). The most important prerequisites for successful postoperative epidural analgesia and thus for increased patient satisfaction are correct selection of the insertion height in relation to the planned operation, constantly available medical pain service, the inclusion of trained care personnel and unequivocal written instructions.