A diverse array of multiwalled carbon nanotube (MWCNT) sensory materials have been synthesized and used to create sensors capable of identifying volatile organic compounds (VOCs) on the basis of their functional groups. Functionalized MWCNTs with a series of cross-sensitive recognition groups were successfully synthesized via zwitterionic and post-transformation synthetic procedures. The incorporated chemical functional groups on MWCNT surfaces introduced greatly increased sensitivity and selectivity to the targeted analytes. The distinct response pattern of each chemical was subjected to statistical treatments, which led to a clear separation and accurate identification of 100% of the VOCs. These results demonstrate that covalent functionalized MWCNT-based sensor arrays are a promising approach for low-cost, real time detection and identification of VOCs.