The sensitivity and specificity of three commercially available complement fixation test (CFT) antigens from c.c.pro (c.c.pro), Central Veterinary Institute of Wageningen UR (CIDC) and the United States Department of Agriculture (USDA) were comparatively evaluated by testing 410 sera collected from glanders-endemic and non-endemic areas (200 true-negative randomly collected sera and 210 sera collected from experimentally immunised animals (12 rabbits, 19 horses), clinically positive (135) and culture-positive (44) horses, donkeys and mules). Immunoblotting (IB) was used as the gold standard test. Highest sensitivity was shown for the CIDC antigen (100 per cent) followed by the c.c.pro antigen (99.39 per cent). However, the USDA antigen showed substantially less (p<0.05) sensitivity (62.19 per cent). Highest specificity was found for the USDA antigen (100 per cent) followed by the CIDC (97.5 per cent) and c.c.pro antigen (96.5 per cent). Positive and negative predictive values (assumed glanders prevalence of <0.1 per cent) for each antigen were calculated to be 95.88 and 99.48 (c.c.pro), 97.04 and 100 (CIDC), 100 and 76.33 per cent (USDA), respectively. Almost perfect agreement (0.96) was found between CFT using either c.c.pro or CIDC and IB.