Intravenous (i.v.) glucocorticoid is recommended for active moderate-to-severe thyroid-associated ophthalmopathy (TAO). However, the details of the treatment schedule are still debatable. The present prospective randomized trial was performed to compare clinical outcomes and serum cytokines between the two regimens. A cohort of 90 patients with active moderate-to-severe TAO was randomized to receive i.v. methyl prednisolone on a weekly protocol or daily scheme. The response rate was evaluated at the 12-week follow-up visit. Serum interleukin (IL)-2, IL-6 and IL-17 levels were measured in 160 patients with TAO, 60 patients with isolated Graves' disease (GD) and 60 normal control (NC) at baseline, as well as patients with active moderate-to-severe TAO at the 12th week after treatment. The daily scheme had a higher response rate than the weekly protocol without a significant difference (77.8 vs. 63.6%, P>0.05). No major adverse events were recorded under either regimen. Overall, minor events were more common on the daily scheme (11.36 vs. 4.35%, P<0.05)than on the weekly protocol, whereas the deterioration of eye symptoms (two patients) was only reported on the weekly protocol. At baseline, the IL-17 level in the TAO group was higher than that in the isolated GD and NC groups (P<0.05). In addition, the IL-17 level in the active TAO group was higher than that in the inactive TAO group (P<0.05). Furthermore, the IL-17 level had significantly decreased under the two regimens at the 12-week visit (P<0.05). In conclusion, for patients with active moderate-to-severe TAO, daily i.v. glucocorticoid therapy has a relative higher response rate than the weekly protocol with a few more minor adverse events. These two regimens have their own merits with regard to adverse effects. IL-17 has the potential to be a biomarker for evaluating TAO activity and treatment effects.