Background Idiopathic granulomatous mastitis (IGM) is a benign disorder of the breast, for which the optimal treatment modality remains missing. Methods A total of 124 patients with a histopathologically proven diagnosis of IGM were enrolled in a prospective, randomized parallel arm study. Patients were treated with topical steroids in Group T (n: 42), systemic steroids (0.8 mg/kg/day peroral) in Group S (n: 42), and combined steroids (0.4 mg/kg/day peroral + topical) in Group C (n: 40). Compliance with the therapy, response to the therapy, the duration of therapy, side effects and the recurrence rates were compared. Results Sixteen patients did not comply with the treatment, and the highest ratio of compliance with therapy was seen in Group T (p < 0.05). Complete clinical regression (CCR) was observed in 90 (83.3%) patients. Response to the treatment (RT) was evaluated radiologically and observed in 89.8% of the patients. There was no statistically significant difference between groups regarding CCR, RT and the recurrence rate. The longest duration of therapy was observed in Group T (22 +/- 9.1-week), whereas the shortest was observed in Group S (11.7 +/- 5.5-week) (p < 0.001). The systemic side effects were significantly lower in Group T in comparison with Groups S and C (2.4% vs. 38.2% and 30.3%, respectively) (p < 0.001). Conclusions The efficiency of the treatment was similar for all groups, both clinically and radiologically. Although the duration of therapy was longer in Group T, the lack of systemic side effects increased the compliance of the patients with the therapy. Therefore, topical steroids would be among first-line treatment options of IGM.