Bacillus species may be resistant to processing and sanitation procedures, making their control an important issue in the food industry. The objective of this study was to develop a rapid method for the differentiation of Bacillus cells at the strain level using infrared microspectroscopy and multivariate pattern recognition techniques. Aliquots (10 ml) of vegetative cells (10 3 CFU/ml) from four strains of each of three Bacillus species (B. cereus, B. mycoides, and B. thuringiensis) were filtered onto hydrophobic grid membranes. The membranes were placed on tryptic soy agar and incubated at 42°C for 24 h and then removed from the agar and dried, and the biomass of individual vegetative colonies was directly measured by attenuated total reflectance infrared (ATR-IR)microspectroscopy. Soft independent modeling of class analogy models generated from second denvative transformed spectra in the 1,300 to 900 cm-1 region exhibited clusters that permitted accurate strain-level classification of all isolates. Major discrimination was related to the signal from phosphate-containing compounds, likely phospholipid. Result; indicate that a simple ATR-IR microspectroscopy technique combined with multivariate analysis could provide the food industry with a rapid and reagent-free screening procedure to complement more elaborate molecular identification methods. Copyright ©, International Association for Food Protection.