Brucellosis research is currently focused on the identification of nonlipopolysaccharide (LPS) antigens which could potentially be useful for the specific serologic diagnosis of brucellosis as well as for vaccinal prophylaxis. On the basis of previous reports, we selected eight Brucella proteins (OMP36, OMP25, OMP19, OMP16, OMP10, p17, p15, and p39) as candidate antigens to be further evaluated. The genes encoding these proteins were cloned, sequenced, and overexpressed in Escherichia coli. The recombinant proteins were purified with a polyhistidine tag and metal chelate affinity chromatography and evaluated in an indirect enzyme-linked immunosorbent assay (iELISA). The specificity of the iELISA was determined with sera from healthy cattle, sheep, and goats and ranged from 95 to 99%, depending on the recombinant antigen and the species tested. Sera from experimentally infected, and from naturally infected, animals were used to evaluate the sensitivity of the iELISA. The antiprotein antibody response was often delayed when compared to the anti-smooth LPS (S-LPS) response and was limited to animals which developed an active brucellosis infection (experimentally infected pregnant animals and sheep and goats from areas where brucellosis is still endemic). Among the recombinant antigens, the three cytoplasmic proteins (p17, p15, and p39) gave the most useful results. More than 80% of the animals positive in S-LPS serology were also positive with one of these cytoplasmic proteins alone or a combination of two of them. None of the recombinant antigens detected experimentally infected nonpregnant cows and sheep or naturally infected cattle. This study is a first step towards the development of a multiprotein diagnostic reagent for brucellosis.