As a potentially blinding disease that can affect 7-27/100000 children, uveitis remains a rare condition, where little evidence based data is available to guide management decisions. Most paediatric uveitis cases are chronic, anterior and non-infectious and may be associated with juvenile idiopathic arthritis (JIA) or remain undifferentiated. Anti-TNF-a treatment is the treatment of choice in patients with uveitis resistant to standard therapy with steroids and traditional immunomodulators (commonly Methotrexate). The aim is a flare free period (quiescence) of 18-24 months, before considering to discontinue any systemic treatment. The aim of this paper was to address the risk of reactivation whilst patients were still on anti-TNF-a (primary outcome), after they stopped it (secondary outcome) and also potential risk factors for reactivation. This retrospective cohort study builds upon a previously published paper by the same authors, demonstrating that 75% of children with non-infectious uveitis achieved quiescence within 12 months with anti-TNF-a (Infliximab or Adalimumab) and that JIA uveitis was most likely to respond to treatment. (J Rheumatol 2013;40:1394-1403). It included patients whose management started between 01/2000 and 07/2012 and casenotes were reviewed, documenting whether there was non-infectious non-traumatic uveitis and whether they had received anti-TNF-a treatment. Quiescence was defined as having ‘slightly active’ (<0.5 cells) or inactive uveitis whilst on <2 drops/day of topical steroids and no oral steroids, sustained for >2 visits over a 28 day period. Infliximab and Adalimumab were the only anti-TNF-a drugs used and any concomitant treatment with steroids, Methotrexate and Mycophenolate Mofetil was documented. Among 50 children who achieved quiescence whilst on anti-TNF-a, 39 were considered at risk for reactivation whilst on anti-TNF-a (primary outcome) and 19 after discontinuation of treatment (secondary outcome). Of that 60% were female, nearly half had JIA and most were treated with Infliximab. The overall proportion relapsing within one year was 27.8% [95% CI]. The risk was higher in children who had discontinued Infliximab at 63.8% versus 21.6% in children still on treatment. In the group that discontinued anti-TNF-a, the risk of reactivation was higher if treated with Adalimumab, rather than Infliximab (HR 13.4, p=0.01). Reactivation was also higher in children that were older at the time of diagnosis of uveitis (HR 1.3, p=0.09). The duration of disease suppression on treatment (quiescence for >1.5 years) had no significant effect on the risk of reactivation. The authors conclude that whilst most children on anti-TNF-a remain quiescent, the majority will reactivate once off treatment. It also seemed that Infliximab was more often followed by remission, but that there was no benefit of prolonged drug induced disease suppression on the reactivation risk. The authors are aware of their study limitations, such as its retrospective nature, the variability in follow-ups and the fact that largely non-uveitis specialists were assessing disease activity. They also point out that some quiescent children may have been lost to follow-up and that there are limitations to the study’s generalisability as the cohort largely consisted of female Caucasians. They recommend further studies to assess whether short versus longer periods of suppression maintenance increase the risk of reactivation after drug cessation, to evaluate the potential advantages of Infliximab versus Adalimumab and to identify possible biomarkers that predict the likelihood of remission after treatment is discontinued.