Halofuginone (Hal) is a coccidiostat used worldwide as a feed additive at a level of 3 ppm in the feed of broiler chickens. The analytical methods described in the literature to determine the levels of Hal in liver are very lengthy and time consuming. This study evaluated the usefulness of determining Hal in chicken serum with a competitive ELISA (cELISA). If a 4‐day withdrawal time could be determined by serum Hal levels, the method would greatly improve on the high‐performance liquid chromatography (HPLC) methods currently used for Hal detection in liver tissue. A modification of a previously developed HPLC method was used to validate the cELISA analysis of Hal in chicken serum. A serum matrix effect that afforded a higher sensitivity by the cELISA for the detection of Hal in chicken serum than in assay buffer or in highly diluted serum was observed. The sensitivity of the cELISA method improved when used in more concentrated serum. The chicken serum samples were evaluated by cELISA using a standard curve obtained in control chicken serum diluted two‐fold with assay buffer. Incurred levels of Hal in broiler chickens fed Hal‐HBr‐treated feed were detected in serum at withdrawal times of 2 and 6 h. At and after 24 h, the residues were not detected by immunoassay with a detection limit of 0.52 ppb or by HPLC with detection limit of 0.86 ppb. The instability of Hal in acidified serum and its potential for methanolysis in the HPLC method were overcome by using the cELISA methodology. Although the determination of Hal in chicken serum by immunoassay is fast, requiring no clean‐up steps, chicken serum cannot be used to determine the required 4‐day withdrawal time in broiler chickens because of the lack of residues in the serum at and after 24 h.