X-ray photoelectron spectroscopy is widely used for the quantitative surface analysis of polymer blends. The compositions of some polymers can change under x-ray exposure, influencing peak intensities and positions in survey spectra and, in addition, the peak shapes in high-energy resolution spectra. In order to quantify spectra acquired after prolonged (>15 min) x-ray exposure, polymer decomposition may need to be taken into account. Calculating the surface composition of a polymer blend using survey and high-resolution core-level spectra is thus not always straightforward.Halogenated compounds, especially chlorinated, are known to degrade with x-ray exposure, resulting in a decrease in the Cl 2p photopeak intensity and an increase in C 1s intensity. In this work, 50 : 50 PVC-PMMA blends are used to evaluate quantitative methods that correct for PVC degradation, based upon either survey or high-resolution spectra. For survey quantification, degradation data were acquired from both blends and pure polymers in order to measure the rate of change of the photoelectron signal intensity. Correction factors were developed based on changes in C 1s, O 1s and Cl 2p intensities as a function of analysis time. High-resolution C 1s spectra were curve-fit using photopeak shapes acquired from PVC and PMMA standards. After degradation, an additional peak attributed to hydrocarbon (HC), which is the product of PVC dehydrochlorination, is required in the curve-fit. Non-uniform HC contamination present initially in some of the PVC-PMMA blends complicates the quantification. The determination of PVC and PMMA concentrations based upon different approaches to the C 1s curve-fitting is presented.The influence of PVC degradation on image appearance and analysis has also been investigated. Principal component analysis (PCA) and classification methods are used to extract maps of the pure components from a degradation image data set. The maps are compared with elemental XPS images and maps of the pure components extracted from an images-to-spectra data set. Analysis of the image data provides insight into the difficulties in basing quantification on the C 1s photopeak curve-fits.