Premedication is not essential to endoscopy but patient tolerance is clearly improved and, thus, ease of examination. Although comparable results can be achieved through nonpharmacologic means, the time and effort involved precludes their widespread use. Despite near universal utilization of premedication in endoscopy, the associated risk is difficult to determine from the available literature. The reported data reveal nominal risk yet must be viewed as minimums. The ideal drug with predictable clinical effects, minimal postprocedure impairment, little respiratory compromise, and proven antagonist is not yet available. Although midazolam seems to represent an advance, recent emphasis on respiratory depression is particularly troublesome. Studies evaluating various agents have suffered from lack of quantitation of such parameters as patient tolerance, ease of examination, and postprocedure impairment. Development of proven standards for these parameters would have to occur before a definitive double-blind randomized trial could be undertaken. Suggested means of assessing these parameters are listed in Table I. Improvement in major morbidity would be difficult in light of its low incidence. As the search for the ideal drug continues, endoscopists must continue to use drugs whose full effects are incompletely understood. The ability to increase patient comfort must be balanced with the small, but ever present, risk of morbidity and mortality.