This study analysed the frequency of distress at induction (DAI) in 2122 paediatric patients. The data were analysed to assess predictors of DAI and to examine associations between predictors of DAI and recovery characteristics. Patient age, preoperative behaviour, premedication (oral midazolam, n = 480) and venue for anaesthesia induction were associated with changes in the incidence of DAI. Distressed preoperative behaviour was a good predictor of DAI in all age groups. Premedication reduced the incidence of DAI in children aged 0.5-2 years old, and in older children who were distressed preoperatively. Induction in the Day Surgery Unit was associated with a reduction of the incidence of DAI in younger children. Children with DAI were more likely to suffer from distress at arousal (P = 0.001). Average early recovery time was prolonged 4.4 minutes and average discharge time in day patients was delayed 36 minutes by the use of oral midazolam premedication. Premedication was not significantly associated with arousal distress. We conclude that a policy of optimizing nonpharmacological approaches for minimizing induction distress, combined with selective premedication with oral midazolam, can produce a low incidence of induction distress and adverse effects.