To investigate the efficacy of IA CS (IAC) therapy in single and multiple joints in children with JIA and to seek for predictors of synovitis flare. The clinical charts of patients who received their first IAC injection between January 2002 and December 2008 were reviewed. The CS used was triamcinolone hexacetonide for large joints and methylprednisolone acetate for small or difficult to access joints. Patients were stratified as follows: one joint injected; two joints injected; and three or more joints injected. Predictors included sex, age at disease onset, JIA category, age and disease duration, ANA status, iridocyclitis, general anaesthesia, number and type of injected joints, acute-phase reactants and concomitant MTX therapy. The cumulative probability of survival without synovitis flare for patients injected in one, two, or three or more joints was 70, 45 and 44%, respectively, at 1 year; 61, 32 and 30%, respectively, at 2 years; and 37, 22 and 19%, respectively, at 3 years. On Cox regression analysis, positive CRP, negative ANA and injection in the ankle were the strongest predictors for synovitis flare. The only significant side effect was skin hypopigmentation or s.c. atrophy, which occurred in <2% of patients. IAC therapy-induced sustained remission of synovitis in a substantial proportion of patients injected either in single or multiple joints, with a good safety profile. The risk of synovitis flare was higher in patients who had positive CRP, negative ANA and were injected in the ankle.