The reproducibility of a commercially available kit utilising enzyme‐linked immunosorbent assay (ELISA) for determination of total aflatoxins has been analysed. Peanut butter samples contaminated with aflatoxin at three concentrations (mean 5.8, 11.7 and 24.8 ng/g) were analysed on different occasions using four kits. Coefficients of variation at all concentrations ranged from 2.6% to 13.6% within replicate, 0.3% to 16.2% within plate and 6.7% to 10.2% between kits. Standard curve precision profiles showed a minimum coefficient variation (CV) of 2–4% and a working range (below 10% CV) covering almost the entire standard curve.