Although the dissociative sedative ketamine is used commonly for pediatric procedural sedation in other settings, the safety of this agent in pediatric gastroenterology is not well-studied. A 5-year experience with ketamine sedation for pediatric gastroenterology procedures was reviewed to document the safety profile of this agent and to identify predictors of laryngospasm during esophagogastroduodenoscopy (EGD). The study was a retrospective consecutive case series of children receiving ketamine administered by pediatric gastroenterologists skilled in basic airway management to facilitate pediatric gastrointestinal procedures during a 5-year period. Patient's records were reviewed to determine indication, dosage, adverse effects, drugs, inadequate sedation, and recovery time for each sedation. A multiple logistic regression analysis was performed to identify predictors of laryngospasm during EGD. Outcome measures were descriptive features of sedation, including adverse effects and predictors of laryngospasm during EGD. During the study period pediatric gastroenterologists administered ketamine 636 times, primarily for EGD (86%) and primarily by the intravenous route (98%). The median loading dose and total dose were 1.00 mg/kg and 1.34 mg/kg, respectively. Inadequate sedation was noted in seven (1.1%) procedures. Adverse effects included transient laryngospasm (8.2%), emesis (4.1%), recovery agitation (2.4%), partial airway obstruction (1.3%), apnea and respiratory depression (0.5%), and excessive salivation (0.3%). There were no adverse outcomes attributable to ketamine. Nearly half (46%) the subjects had severe underlying illness (American Society of Anesthesiologists (ASA] class > or =3). All instances of laryngospasm occurred during EGD (9.5% incidence), and the only independent predictor of laryngospasm in this sample was decreasing age. The incidence of laryngospasm was 13.9% in preschool-aged (< or =6 years) children and was 3.6% in school-aged (>6 years) children (difference 10.3%, 95% confidence intervals 5.5-14.9%). No dose relationship was noted with laryngospasm, and the risk did not increase with underlying illness. Pediatric gastroenterologists skilled in ketamine administration and basic airway management can effectively administer this drug to facilitate gastrointestinal procedures. Transient laryngospasm occurred in 9.5% of children receiving ketamine for EGD, and its incidence was greater in preschool than in school-aged children.