The long-term effect of prednisone in Crohn's disease has been examined in a double-blind controlled trial. Clinical relapse, recurrence, and extension of the disease were examined in 64 patients followed-up for up to three years. Fourteen patients were withdrawn because of severe symptoms (eight on prednisone and six controls); the withdrawal rate in both groups was 30% at three years. Nine other patients had radiological recurrence or extension of disease (five prednisone and four controls). Prednisone did not improve the relapse rate, nor did it affect recurrence or extension of disease.