Results from two 1980 experiments were used to compare the performance of the visual, minicolumn and thin-layer chromatography (TLC) methods in detecting loads of farmers stock peanuts with aflatoxin. The first experiment was conducted to establish variability of the methods under ideal conditions. The second experiment was conducted to evaluate the performance of the three methods under commercial conditions. In this paper, data and empirical models were used to explore potential improvements in detecting low-level contaminated lots. The minicolumn and TLC methods were very sensitive and provided consistent measurements. Operation characteristic (OC) curves developed by using empirical models (Logistic and Gompertz) were compared to statistical distribution functions used by Whitaker and coworkers. Assuming a desired probability of acceptance of 15%, both the minicolumn and TLC methods would accept lots with 60 ppb aflatoxin, while the visual methods would accept lots with 150 ppb aflatoxin. In crop years similar to 1980, the currently used method would not segregate peanuts to meet the 1–5 ppb (total) tolerance levels and would provide excessive sheller risk. Because of their objectivity, precision and higher sensitivity, the minicolumn and TLC have potential for improving the detection of contaminated loads of farmers stock peanuts.