The intravenous methylprednisolone (iv MP) strategy for Graves' ophthalmopathy (GO) and evaluation of its activity against the disease warrants further exploration. A prospective randomized controlled trial for 3 months was performed in a tertiary referral teaching hospital to compare the efficacy and safety of two different regimens of iv MP, and determine the value of clinical activity score (CAS) and T2 relaxation time (T2RTs) and areas of extraocular muscles (EOMs) by magnetic resonance imaging for diagnosis of active GO. Forty patients with moderate to severe GO and CAS ≥ 3 or 1 ≤ CAS < 3 with prolonged T2RTs on EOMs were randomly assigned to a monthly (MR: 1.5 g iv MP monthly for 3 months) or weekly (WR: 0.5 g iv MP weekly for 6 weeks, followed by 0.25 g weekly for 6 weeks) regimen. Overall response based on ophthalmic symptoms, T2RTs, areas of EOMs and adverse effects were recorded at each visit. The total rate of response was 71.9%. Rates of improved, unchanged, deteriorated were similar between the MR and WR groups (p>0.05). The maximum T2RTs and areas significantly decreased at the end of intervention in both groups (p<0.05). Results show that both MR and WR are effective and safe in treatment of GO. T2RTs combined with CAS can sensitively detect active GO and predict the response to iv MP.