Survival rates of two groups of corticosteroid-treated patients with systemic lupus erythematosus (SLE) have been compared using life-table analysis. Group 1 consists of 209 SLE patients studied in New York City between 1957 and 1968; group 2 consists of 156 SLE patients studied in Connecticut between 1968 and 1976. A similar percentage of patients in each group was followed up closely by one of us. The five-year survival rate for group 1 was 70%, compared with 93% for group 2. The improved survival could not be explained by differences in racial composition, incidence of CNS manifestations, or incidence of severe renal disease. The markedly improved survival may be due to several factors, including a generally better understanding of the disease and the recent use of newer antibiotics. The management of the two groups was similar, except that group 2 patient's conditions were managed not only on the basis of clinical evidence of disease activity but also on the basis of the serum complement (C3) level and on the basis of the level of antibody to native DNA. The data suggest that the use of such serologic values may lead to improved survival.