In order to study the false positive problem during detecting heme binding protein with horseradish peroxidase (HRP) as secondary antibody, spinach cytochrome [WTBX]b₆f[WTBZ] complex (Cyt [WTBX]b₆f[WTBZ]) and ferredoxin-NADP⁺ oxidoreductase (FNR) are used as the test material (the former contains the heme binding protein Cyt f, and the latter does not contain heme group). A comparative experiment is designed through the use of the primary antibody of spinach FNR and secondary antibodies labeled with HRP and alkaline phosphatase (ALP). The results indicate that in the HRP-labeled secondary antibody system, Cyt f could be colored even without the incubation by the primary antibody and the secondary antibody, which is a significant false positive. In the ALP-labeled secondary antibody system, the Cyt f doesn't show color, and only the FNR is colored. In western blot experiments, if the protein to be detected is a heme-binding protein or contains a small amount of heme-binding protein, HRP-labeled secondary antibodies should be avoided and ALP-labeled secondary antibodies or other secondary antibodies should be recommended. It is speculated that the heme-binding protein has an HRP-like oxidoreductase activity, namely it can react with the color developing solution, thereby generating a false positive and interfering with the judgment of the experimental results.