The diagnosis of brucellosis is frequently a difficult one, and it is founded basically on the results of blood cultures and serological tests. Wright's classical agglutination test is commonly utilized in the serological diagnosis of brucellosis and its value is well established in the acute forms of the disease; for the diagnosis of subacute or chronic forms, however, a Coombs test must be performed to show incomplete antibodies. The agglutination carried out with serum treated with 2-mercaptoethanol allows the determination of the type of immunoglobulins present. Indirect immunofluorescence has been infrequently utilized in the diagnosis of brucellosis. The first three tests have been compared with indirect immunofluorescence using sera from 100 patients presenting with fever of unknown origin. All tests were negative in 40 sera, and of the 60 remaining sera only 32 were positive to all four tests. The highest titers were found with the Coombs test, while immunofluorescence yielded titers which were never above 1/320. Significant correlation coefficients were obtained between the tests. The best correlation was obtained between immunofluorescence and 2-mercaptoethanol, followed by immunofluorescence and Coombs, and by immunofluorescence and agglutination. Only immunofluorescence titers above 1/80 corresponded to the same of superior titers in the remaining tests; thus immunofluorescence titers below 1/80 can not be considered as diagnostic of the disease. The authors believe that immunofluorescence to 1/80 can be utilized as a quick routine test for the diagnosis of brucellosis, its disadvantages being a high cost and the requirement for an immunofluorescence microscope which may not be available in many laboratories.