A review of 28 patients with nonthymomatous myasthenia gravis who underwent thymectomy at the University of Colorado Health Sciences Center, Denver, from 1967 to 1979 shows significant stepwise changes in management and results. Comparison among three periods--period 1 (1967 to 1971), when thymectomy competed with prednisone,which were not given in the perioperative period (seven patients); period 2 (1974 to 1976), when thymectomy was followed by six months of prednisone therapy (ten patients); and period 3 (1977 to 1979), when prednisone was also given to prepare the patients for thymectomy (11 patients)--demonstrated a decreasing need for tracheostomy and respiratory support (86% v 10% v 0%), shorter stay in the intensive care unit (21 v 3 v 1 day), and shorter hospitalization (36 v 13 v 4 days). Remission or marked amelioration of symptoms occurred in 56% of group 1 and 100% of both groups 2 and 3. Earlier application of thymectomy and its performance through a short upper transverse sternotomy incision also contributed to the improved results.