An investigation into the rapid detection of mycotoxin-producing fungi on corn by two mid-infrared spectroscopic techniques was undertaken. Corn samples from a single genotype (RWA2, blanks, and contaminated with Fusarium graminearum) were ground, sieved and, after appropriate sample preparation, subjected to mid-infrared spectroscopy using two different accessories (diffuse reflection and attenuated total reflection). The measured spectra were evaluated with principal component analysis (PCA) and the blank and contaminated samples were classified by cluster analysis. Reference data for fungal metabolites were obtained with conventional methods. After extraction and clean-up, each sample was analyzed for the toxin deoxynivalenol (DON) by gas chromatography with electron capture detection (GC-ECD) and ergosterol (a parameter for the total fungal biomass) by high-performance liquid chromatography with diode array detection (HPLC-DAD). The concentration ranges for contaminated samples were 880-3600 microg/kg for ergosterol and 300-2600 microg/kg for DON. Classification efficiency was 100% for ATR spectra. DR spectra did not show as obvious a clustering of contaminated and blank samples. Results and trends were also observed in single spectra plots. Quantification using a PLS1 regression algorithm showed good correlation with DON reference data, but a rather high standard error of prediction (SEP) with 600 microg/kg (DR) and 490 microg/kg (ATR), respectively, for ergosterol. Comparing measurement procedures and results showed advantages for the ATR technique, mainly owing to its ease of use and the easier interpretation of results that were better with respect to classification and quantification.