Although the concept of continuous infusion (CI) of factor concentrates is well known, prospective paediatric data are rare. We present a prospective open-labelled non-randomized study focusing on safety, efficacy and factor VIII (FVIII) usage compared with bolus injections (BI) in children. In 43 consecutive patients (0.5–17 years; median: 9.6) undergoing different operations, CI was started with an initial FVIII-bolus of 70 IU kg−1 bodyweight, followed by a median infusion rate of 4.4 IU kg−1 h−1 (range: 2.8–9.5) dose adjusted for daily FVIII levels (target: 60–80%). No direct serious adverse events occurred; however, two out of 43 patients, both from the group of four patients with less than 20 exposure days (ED) before starting CI, developed a high-responding inhibitor. Two CI patients showed mild thrombophlebitis or rash. Infusion rates needed to achieve adequate FVIII levels were highly predictable and could be reduced because of decreasing FVIII clearance. Bleeding, requiring additional boli, was observed in eight out of 43 patients. Therapy duration and factor usage of CI were influenced by the procedure, but not by the product used or thrombophilia. Twelve of these CI patients were compared with 12 contemporary consecutive age- and procedure-matched BI patients. Compared with BI patients, CI patients saved 30% FVIII (812.9 vs. 563.2 IU kg−1, P < 0.006). We conclude that CI forms a safe and effective method for perioperative care in children and reduces factor usage. Because of the unknown risk of inhibitor development, we will use CI only in patients beyond 20 ED.