A prospective, randomized and double blind study was undertaken to determine the incidence and a possible dose- or age-dependence of hiccups in children premedicated with rectal midazolam and to investigate the treatment of hiccups by intranasal ethyl chloride spray application. Two hundred ASA physical status 1 and 2 children, weighing 3.0 to 15.0 kg, scheduled for minor surgery, were randomly assigned to be given either 0.5 mg.kg-1 midazolam(n=100) or 1.0 mg. kg-1 midazolam (n=100) administered rectally. If hiccups were observed during a period of 20 min after premedication with midazolam, these children were treated after 3 min of hiccups with two short intranasal applications of ethyl chloride spray. Hiccups occurred in 22% of children in the 0.5 mg.kg-1 group and 26% in the 1.0 mg.kg-1 group (n.s.). The intranasal application with ethyl chloride was successful in 100% in both groups. The mean age levels between children with or without hiccups were 5+/-9 months vs 21+/-19 months (P<0.01) in the 0.5 mg.kg-1 group and 6+/-7 months vs 20+/-14 months (P<0.01) in the 1.0 mg.kg-1 group. Intranasal application of ethyl chloride spray seems to be an effective therapy for midazolam induced hiccups in paediatric anaesthesia. The incidence of these hiccups is highly age significant, but not dose dependent.