Abstract Objectives To assess the efficacy and tolerance of the conventional first-line treatment by MTX and CS in patients with JDM regardless of severity. Methods We conducted a monocentric retrospective study of patients with newly diagnosed JDM treated with MTX and CS from 2012 to 2020. The proportion of clinically inactive disease (CID) within 6 months of MTX initiation was evaluated using both Paediatric Rheumatology International Trials Organisation (PRINTO) criteria (evaluating muscle inactive disease) and DAS (evaluating skin inactive disease). We compared responders and non-responders using univariate analyses. Results Forty-five patients with JDM, out of which 30 (67%) severe JDM, were included. After 6 months of treatment with MTX and CS, complete CID, muscle CID and skin CID were achieved in 14/45 (31%), 19/45 (42%) and 15/45 (33%) patients, respectively. The absence of myositis-specific (MSA) or myositis-associated autoantibodies (MAA) at diagnosis was associated with a better overall, cutaneous and muscular therapeutic response, compared with antibody-positive forms (P < 0.01). Requirement for ICU (P = 0.029) and cutaneous ulcerations (P = 0.018) were associated with a less favourable muscle response. MTX was stopped due to intolerance in six patients (13%) before month 6. Conclusions Conventional first-line treatment with MTX was not efficient in a large subset of JDM patients, especially in patients with MSA-positive forms, and in patients with severe JDM. Larger, multicentre cohorts are required to confirm these data and to identify new predictive biomarkers of MTX response, in order to treat patients with JDM as early as possible with appropriate targeted drugs.