Medications are routinely administered to children prior to upper gastrointestinal endoscopy procedures. We evaluated the efficacy and safety of four dosage regimens: meperidine 2 mg/kg (M); low-dose meperidine 1 mg/kg and diazepam 0.1 mg/kg (M low+D); high-dose meperidine 2 mg/kg and diazepam 0.1 mg/kg (M high+D); and, diazepam 0.1 mg/kg (D) in a randomized, double-blind trial in 71 pediatric patients (ages 1 to 19 years). Cooperation, emotional state, and sedation were rated. The frequency of negative behavior demonstration (i.e., crying, gagging, flailing, nervous behavior) indicating ineffective or inadequate sedation was recorded before, during, and after the procedure. Vital signs and oxygen saturation were noted. Significant differences were observed in the efficacy of the various sedation regimens. Overall, the physicians and nurses rated M most effective in children < 11 years. In these children, M high + D and D were rated least effective by the nurses, whereas physicians felt that D was least effective. In the older children, M, M low + D, and M high + D were rated similarly effective by physicians and nurses, and D was rated least effective. In both age groups, the lowest incidence of negative behaviors during the endoscopy procedure occurred in patients who received M. Combinations of D and M + D resulted in an increased incidence of negative behaviors; thus, M may be the preferred sedation regimen in children undergoing endoscopy. The addition of diazepam to meperidine may be detrimental in terms of eliciting negative behaviors.