In this study, 241 serum samples from individuals exposed to brucellosis were subjected to the Rose Bengal plate test (RBPT); the titre was estimated by standard tube agglutination test (STAT), with positive > or = 80 IU/ml. Randomly selected sera (n = 81) were analysed by complement fixation test (CFT): titre > or = 1:4 was considered positive. Of 241 sera subjected to RBPT and STAT, 177 were negative in both tests; 5 samples tested negative by RBPT but positive by STAT. None was positive by RBPT and negative by STAT. Of 81 sera subjected to CFT, 23 (28.4%) were positive. Both RBPT and CFT found 18 samples positive; 5 samples were positive by CFT and negative by RBPT. Comparison of STAT with CFT showed 13 samples positive by STAT but negative by CFT, and 4 positive by CFT but negative by STAT. The sensitivity and specificity of STAT were 82.6% and 77.6%, respectively, with CFT as gold standard. No test is perfect, and the clinical history coupled with a combination of two or more tests will reduce diagnostic errors.