OBJECTIVE To describe the incidence and intensity of pain, hypotension and other epidural-related side-effects after major abdominal surgery and to identify factors associated with effective epidural analgesia. BACKGROUND Evidence exists that up to 30% of patients with epidural analgesia still experience severe pain in clinical practice. When epidurals produce good pain relief, potentially harmful side effects can result. METHODS Data were collected prospectively from 480 consecutive general surgical patients in a large District General Hospital in the UK. RESULTS Fifty-six per cent of patients were hypotensive on day one. Low pain scores were found to be a predicator for postoperative hypotension. Severe pain was associated with emergency patients, male gender, the absence of hypotension and an epidural that did not continue until planned removal. Significantly lower mean pain scores were found in patients who were 70 years of age and over. No correlation exists between chronic pain before surgery and the level of postoperative pain. CONCLUSIONS This study contributes to the epidural versus standard analgesia debate by describing the population in whom epidurals are used and identifies factors associated with both the success and failure of the technique in everyday practice. The Audit Commission has proposed a standard whereby less than 5% of patients should suffer severe pain following surgery. This may prove to be an unrealistic goal with currently available techniques. Relevance to clinical practice.  As it is ward nursing staff who are primarily responsible for monitoring the effectiveness of the technique, it is important for nursing professionals to know the potential risks of postoperative epidural analgesia to be able to respond appropriately. The important measurements of an optimum acute pain assessment have been identified, based on the study results, allowing the development of a national acute pain registry to inform future practice.