Acarus siro L. 1758 (Acari: Acaridida: Acarididae) is an important pest of stored grain because it contaminates the grain by allergens and transfers pathogenous microorganisms. Rapid detection of contamination enables to intercept an early grain infestation by the pest. In this study, we compared the usability and efficiency of various detection approaches. Under laboratory conditions, grain samples of various sizes were infested by different levels of the following contaminants: eggs, adults, and feces of A. siro. The samples were analyzed by enzyme-linked immunosorbent assay (ELISA) by using anti-A. siro polyclonal antibody (immunochemical method), extracted in Berlese-Tullgren funnels, sieved, and processed by filth-flotation (conventional methods). The adults or juveniles of A. siro could be detected by all the three tested conventional methods and ELISA with detection limits in the range from 221 to 1,157 mites/kg grain. Eggs were detected by filth-flotation only; the detection limit was 1,950 eggs/kg grain. The feces of A. siro were detectable by ELISA test, only. ELISA enabled the detection of the feces with the minimal threshold level of 1.04 μg feces/g grain; it means the assay allowed to trace less than one metabolically active mite per gram of grain. The study thus demonstrated that reliable A. siro detection in grain can only be achieved by combining different detection methods. European Union and U.S. administratives dictate zero or near-zero tolerance level for mite infestation in stored products. This demand is difficult to fulfill, because every detection method is limited by its detection limit; thus, it is hard to reliably detect infestation levels lower than obtained detection limits. This methodical limitation is discussed in context with the determined detection limits of the tested methods.