A carbon nanotube based gas sensor array was developed for discriminating gases and vapors. The sensor array was composed of 32 sensing elements with nanomaterials, e.g. pristine single walled carbon nanotubes (SWNTs), and SWNTs with different metal dopants and polymer coatings. This sensor array was exposed to NO2, HCN, HCl, Cl-2, acetone and benzene in parts per million (ppm) concentration levels. The array data was normalized and autoscaled for eliminating concentration and background noise, and ignoring outliers. The post processed array data was then subjected to a principal component analysis, a pattern recognition technique, for gas and vapor discrimination. All tested gases and vapors can be discriminated by their chemical nature in the low gas/vapor concentration at ppm levels using this carbon nanotube based sensor array. (c) 2006 Elsevier B.V. All rights reserved.