Arrays of broadly cross-reactive vapor sensors provide a man-made implementation of an olfactory system, in which an analyte elicits a response from many receptors and each receptor responds to a variety of analytes. Pattern recognition methods are then used to detect analytes based on the collective response of the sensor array. With the use of this architecture, arrays of chemically sensitive resistors made from composites of conductors and insulating organic polymers have been shown to robustly classify, identify, and quantify a diverse collection of organic vapors, even though no individual sensor responds selectively to a particular analyte. The properties and functioning of these arrays are inspired by advances in the understanding of biological olfaction, and in turn, evaluation of the performance of the man-made array provides suggestions regarding some of the fundamental odor detection principles of the mammalian olfactory system.