Mycotoxins represent an heterogeneous group of toxic compounds from fungi metabolism. Due to the frequent occurrence of mycotoxins in cereals commodities the develop of cost/effective screening methods represent an important topic to ensure food and feed safety. In the presented study a commercial electronic nose constituted by ten MOS (Metal Oxide Sensors) was applied to verify the possibility of discriminating between mycotoxins contaminated and non-contaminated cereals. The described analytical approach was able to discriminate contaminated and non-contaminated samples both in the case of aflatoxins infected maize and deoxynivalenol infected durum wheat samples. In the case of maize data two sensors from the array revealed a partial relation with the level of aflatoxins. These results could be promising for a further improvement of electronic nose application in order to develop a semi-quantitative screening approach to mycotoxins contamination.