Oral prednisone is considered the standard first-line therapy of adult immune thrombocytopenia, but its long-term efficacy is limited. We performed a prospective, randomized, multicenter trial comparing daily prednisone (1-2 mg/kg/day for 2-4 weeks with subsequent dose reduction) with six 3-week cycles of pulsed dexamethasone (0.6 mg/kg/day, days 1-4). The primary endpoint was remission duration. Of 26 patients enrolled, 22 were evaluable for response. Nine were treated with prednisone and 13 with dexamethasone. The median follow-up was 46 months. The initial response rate (PLT ≥50 × 109/l) was 100% in both groups. Long-term remissions were significantly more frequent with pulsed dexamethasone than with daily prednisone (12 months posttreatment: 77 vs. 22%; p = 0.027). The side effects were similar, but patients on dexamethasone suffered significantly more often from insomnia, while patients on prednisone tended to have more infectious complications. Although the cumulative cortisol equivalent dose was comparable during the first 4 weeks of therapy, it was significantly higher in the dexamethasone arm than in the prednisone arm during the ensuing treatment period. We conclude that repeated cycles of pulsed dexamethasone are a good alternative to daily prednisone as a first-line treatment of immune thrombocytopenia. The duration and intensity of glucocorticoid therapy are important determinants of treatment outcome.