Internet-delivered cognitive behaviour therapy (ICBT) is emerging as a powerful tool to fill the gap between demand and availability of evidence-based treatment for paediatric anxiety disorders. However, it is still unclear how to best implement it in routine clinical care. 123 children (8-12 years) with anxiety disorders underwent a 12-week ICBT programme with limited therapist support. Participants were assessed 3- and 12-month post-ICBT (3MFU and 12MFU, respectively). Non-remitters who still fulfilled diagnostic criteria for their principal anxiety disorder at 3MFU were offered additional manualised face-to-face (F2F) CBT. The aim of the study was to emulate a stepped-care model of health care delivery, where the long-term treatment gains of ICBT as well as the potential benefit of proving addition treatment to non-remitters of ICBT were evaluated. Remitters of ICBT (n = 73) continued to improve throughout the study period (pre-ICBT to 12MFU; Cohen'sd = 2.42). At 12MFU, 89% (n = 65) were free from their principal anxiety disorder. Of all the participants classed as non-remitters at 3MFU (n = 37), 48.6% (n = 18) accepted the offer to receive additional F2F CBT. These participants also improved with a large effect from pre-ICBT to 12MFU (Cohen'sd = 2.27), with the largest effect occurring during F2F CBT. At 12MFU, 83% (n = 15) were free from their principal anxiety disorders. The majority of non-remitters declining additional F2F CBT (63.2%;n = 12) did so due to already receiving treatment at their local CAMHS, prior to 3MFU. The effects of ICBT for anxiety disorders are durable at least up to 1 year after the end of treatment. Patients who fail to fully benefit from ICBT improved further with additional F2F sessions at our clinic, suggesting that it may be feasible to implement ICBT within a stepped-care model of health care delivery.