A multivariate curve resolution (MCR) method that combines evolving factor analysis (EFA) with alternating least squares (ALS) is applied to partially co-eluting vapors measured with a thiolate-monolayer-protected gold nanoparticle (MPN)-coated chemiresistor array used as the detector in a microfabricated gas chromatograph (µGC). Two pairs of vapors having different array response pattern similarities (ρ) were selected as analytes: trichloroethylene (TCE) and n-heptane (HEP), which have relatively similar patterns (ρ=0.80), and cyclohexane (CHX) and n-butanol (BOH), which have relatively dissimilar patterns (ρ=0.20). Calibration curves and a response pattern library for the individual vapors were established before the analysis. Binary mixtures were generated and analyzed at various values of chromatographic resolution (R = 0.1 to 1.0) and relative concentrations corresponding to relative response ratios ranging from 1:10 to 10:1 for the least sensitive sensor in the array. EFA-ALS analysis permitted extraction and reconstitution of responses of the individual components of the binary composite peaks with high fidelity (r > 0.95 in all cases for TCE+HEP and > 0.9 in most cases with CHX+BOH). Subsequent pattern matching with response patterns in the library was successful in all cases. Quantification of recovered responses is being explored. This project was funded by the the Department of Defense, Environmental Security Technology Certification Program (DoD-ESTCP).