The use of Fourier transform near-infrared (FT-NIR) spectroscopy and multivariate pattern recognition techniques for the rapid detection and identification of bacterial contamination in liquids was evaluated. The complex biochemical composition of bacteria yields FT-NIR vibrational transitions (overtone and combination bands) that can be used for classification and identification. Bacterial suspensions (Escherichia coli HB101, E. coli ATCC 43888, E. coli 1224, Bacillus amyloliquifaciens, Pseudomonas aeruginosa, Bacillus cereus, and Listeria innocua) were filtered to harvest the cells and eliminate the matrix, which has a strong NIR signal. FT-NIR measurements were done using a diffuse reflection-integrating sphere. Principal component analysis showed tight clustering of the bacterial strains at the information-rich spectral region of 6000-4000 cm-1. The method reproducibly distinguished between different E. coli isolates and conclusively identified the relationship between a new isolate and one of the test species. This methodology may allow for the rapid assessment of potential bacterial contamination in liquids with minimal sample preparation.