The use of classic methods of diagnosis of brucellosis was analyzed, particularly serologic methods whose use in endemic areas and risk groups has been questioned in the literature.Prospective analysis of these methods was performed in a group of 171 patients suspected as having brucellosis proceeding from an endemic area, with progressions of risk and frequent antecedents of brucellosis. The results obtained were compared in 119 patients in whom brucellosis was confirmed (80 with positive cultures and 39 with clinic-serologic diagnosis) and in 52 in whom the diagnosis was excluded.The hemocultures provided a sensitivity of 70% with a mean delay in growth of 13.6 days. The rise of Bengal test showed sensitivity of 95% and specificity of 75%. The most adequate cut-off point for seroagglutination was of 1/160 and for the Coombs test 1/320 with sensitivities of 93 and 92% and specificities of 97 and 100%, respectively.The routine serologic tests offer good results for the diagnosis of brucellosis in endemic areas upon use of adequate cut-off points and permitting therapeutic decisions to be taken prior to knowledge of the results of the cultures. The rose of Bengal test is valid for initial selection of this type of population.