Comparison of two commercial methods with a UHPLC–MS/MS method for the determination of multiple mycotoxins in cereals

•High false-negative rates of mycotoxin test strips call for alarm.•Results of ELISA kit is inaccurate within its claimed detection range.•The repeatability of most ELISA kits at the declared LOD is unacceptable.•Derivatives of deoxynivalenol and zearalenone can cause cross-reactivity.•Sample matrix and cumulative exposure of derivatives can amplify cross-reactivity. Immunoassay-based techniques are important on-site screening tools for the detection of mycotoxins in cereals. This study aims to evaluate the trueness, precision, repeatability and cross-reactivity of commercially available test strips, ELISA kits and UHPLC-MS/MS on analyzing zearalenone, ochratoxin A, deoxynivalenol, T-2 toxin and fumonisin B1. The results showed that false negative rate (25.7 %–37.4 %) of all tested mycotoxins by test strips was higher than the false positive rate (0 %–31.0 %). The repeatability of ELISA kits at the declared LOD dispersed from −85.7 % to +98.4 %. ELISA kits were more accurate at 50 % of the maximum residue limit (MRL) of mycotoxins than 150 % and 200 %. All the tested deoxynivalenol/zearalenone derivatives had cross-reactivity with different level, and sample matrix could reinforce this overestimation of target mycotoxin. This study emphasized that higher-quality antibody screening and more analytical performance investigations are need to address for on-site detection of mycotoxins in the future.