Heavy i.v. sedation is often used in upper GI endoscopy. Sedation, however, creates the need for recovery facilities and precludes patients from returning to their normal daily activities. This is undesirable, since endoscopy is routinely performed as an out-patient procedure. Also the cost for medication and recovery facilities militate against the indiscriminate use of i.v. sedative premedication. The present study was undertaken in an attempt to establish what proportion of patients can benefit from oral premedication, and whether such an administration route can eliminate some of the disadvantages associated with i.v. sedation. Four hundred out-patients were randomized to receive orally either triazolam, 0.125 mg, or placebo. Of the patients, 359 were evaluable; 177 received placebo and 182 triazolam. All major aspects of the procedure were covered using visual analogue scale questionnaires for the endoscopist and patient. There were no differences in endoscopic experience, or sex and age distribution between the groups. Triazolam reduced patient discomfort, 38.6 +/- 25.6 vs 44.8 +/- 30.1 (p = 0.0379). Recollection of post-endoscopy information was the same in both groups. One patient complained of drowsiness following the procedure. No patient needed to stay in hospital to complete recovery. Endoscopy quality was identical in the two groups. Oral premedication has the potential to be of significant value, may optimize the use of endoscopy resources, and does not impair patient activities post-endoscopy.