Including the tumor necrosis factor TNF inhibitors (TNFi) in the management guidelines for children with juvenile non-systemic juvenile idiopathic arthritis (JIA) has significantly increased the effectiveness of antirheumatic treatment. However, the guidelines of TNFa withdrawal when disease remission is achieved remains unclear. Objective. To study the efficacy and safety of different TNFi withdrawal regimens in patients with juvenile idiopathic arthritis without systemic manifestations due to the achievement of long-term disease remission. Patients and methods: The prospective simple randomized trial included 76 patients with JIA without non-systemic juvenile idiopathic arthritis, with a disease remission duration of ≥24 months, developed under the conditions of treatment with TNFi. Of these, 38.2% are male, the average age is 11.6 years; 67 (88.2%) patients received etanercept (0.8 mg / kg / week. subcutaneously 1 time per week), the rest – adalimumab (24 mg/m2 subcutaneously 1 time in 2 weeks). Depending on the withdrawal regimen, patients were divided into three groups: in the I group, one-time withdrawa was presented, in the II group, the interval between administrations was of two times and the III group was reduced by a dose of two times. The main indicator of the study: preservation of the stage of inactive disease / remission (according to the criteria for remission of C. Wallace) for 6, 12 and 18 months after the withdrawal of TNFi. Results. At the time of withdrawal of TNFi, the average duration of the disease was 7.3 years, the duration of therapy with TNFa was 5 years, remission – was 4.4 years. For 6, 12 and 18 months after the withdrawal of TNFi, the stage of remission persisted in 44 (57.9%), 21 (27.6%) and 18 (23.7%) patients. Exacerbation of disease was registered in 58/76 (76%) patients. Biologic therapy was resumed in 56/58 children, in all cases the stage of inactive disease was reached. Among patients with exacerbation of JIA, 23/56 (41%) continued to receive disease-modifying antirheumatic drugs. Predictors of successful withdrawal of TNFa in patients with non-systemic JIA 6 and 12 months after discontinuation of the biological therapy were the female sex, the absence of HLA B27, antinuclear factor and long-term methotrexate therapy; predictors of exacerbation - increasing in the serum concentration of calprotectin (S-100 protein), the presence of subclinical synovitis according to ultrasound and MRI of the joints and uveitis. Conclusion. After the withdrawal of TNFi, the remission stage for 6 and 12 months persisted more than in half and one third of patients. The withdrawal regimens did not affect the duration of the remission of disease. The resumption of biological therapy in the exacerbation stage of JIA made it possible to reach the remission stage in most patients. Predictors of successful withdrawal of TNFi in patients with non-systemic JIA 6 and 12 months after discontinuation of the biological therapy were the female sex, the absence of HLA B27, antinuclear factor and long-term methotrexate therapy; predictors of exacerbation – increasing in the serum concentration of calprotectin (S-100 protein), the presence of subclinical synovitis according to ultrasound and MRI of the joints and uveitis. The presence of exacerbation predictors makes it impractical the withdrawal of biological therapy due to the high risk of developing an exacerbation of disease. Key words: (TNF)-alpha inhibitors, juvenile idiopathic arthritis, long-term remission, withdrawal predictors, high-sensitivity C-reactive protein (hsCRP), serum calprotectin (S-100 protein)