This randomized, double-blind, placebo-controlled study compared the efficacy of inhaled nitrous oxide (N(2)O) with enteral midazolam for sedation of children with cerebral palsy (CP) undergoing botulinum toxin A (BoNT-A) injections. Fifty children (29 males, 21 females; mean age 8y 2mo [SD 4y 5mo]; range 1-16y) were randomized to sedation with N(2)O (n=25) or midazolam (n=25). Groups were similar in type of CP (diplegia, 11; triplegia, three; quadriplegia, 16; hemiplegia, 16; other, three) and Gross Motor Function Classification System level (Level I, 4; II, 24; III, 4; IV, 13; V, 5). Both groups were equally sedated at time of injection (p=0.661), but those in the midazolam group were more sedated at time of discharge (p<0.001). N(2)O was more effective in reducing pain compared with midazolam as measured using the Face, Legs, Activity, Cry, Consolability (FLACC) scale (p=0.010), parental estimate of pain (p=0.009), and nursing estimate of pain (p=0.007). Parents in the N(2)O group rated it better than prior sedation with midazolam for BoNT-A injections (p=0.031). Physicians and nurses reported no difference in ease of procedure between the groups. One child in the midazolam group and eight in the N(2)O group had adverse effects, all of which resolved promptly. N(2)O appears to be an effective means of sedation for children undergoing outpatient BoNT-A injections.