Objectives: To determine whether midazolam, when used as an induction agent for emergency department (ED) rapid‐sequence intubation (RSI), is used in adequate and recommended induction doses (0.1 to 0.3 mg/kg), and to compare the accuracy of the dosing of midazolam for ED RSI with the accuracy of dosing of other agents. Methods: The authors conducted a systematic query of a prospectively collected database of ED intubations using the National Emergency Airway Registry data, gathered in 11 participating EDs over a 16‐month period. A data form completed at the time of emergency department intubation (EDI) enabled analysis of patients' ages, weights, and indications for EDI, as well as the techniques and drugs used to facilitate EDI. Data were analyzed to determine whether midazolam is used in recommended doses during RSI. Patients intubated with midazolam alone were compared with patients who received other induction agents for RSI. Results: Of 1,288 patients entered in the study, 1,023 (79%) underwent RSI. Of the 888 RSI patients with an age recorded, midazolam was used as the sole induction agent in 140 (16%). The mean (±SD) dosages of midazolam used in RSI were 2.6 (±1.7) mg in children (age ≤ 18) and 3.7 (±2.5) mg in adults (age ≥19); the mean (±SD) dosages by weight were 0.08 (±0.04) mg/kg in children and 0.05 (±0.03) mg/kg in adults. More than half (56%) of the children, and nearly all (92%) of the adults, received dosages lower than the minimum recommended dosage (0.1 mg/kg). Of patients who received barbiturates, only 21% of children and 21% of adults received a dose lower than the minimum recommended. When combined with another induction agent, midazolam was dosed similarly to when it was used alone: mean adult doses were 3.1 (±1.2) mg and 0.04 (±0.02) mg/kg. Conclusions: Underdosing of midazolam during ED RSI is frequent, and appears to be related to incorrect dosage selection, rather than to a deliberate intention to reduce the dose used.