Field asymmetric waveform ion mobility spectrometry (FAIMS) has significant potential for post-ionization separations in conjunction with MS analyses. FAIMS fractionates ion mixtures by exploiting the fact that ion mobilities in gases depend on the electric field in a manner specific to each ion. Nearly all previous work has used pure gases, for which FAIMS fundamentals are understood reasonably well; however, unexpected phenomena observed in some gas mixtures (e.g., N(2)/CO(2)) but not in others (N(2)/O(2)) remain unexplained. Here, we introduce and experimentally test a universal model for FAIMS separations in mixtures, derived from formalisms that determine high-field mobilities in heteromolecular gases. Overall, the theoretical findings are consistent with data for N(2)/CO(2) (although quantitative discrepancies remain), while results for N(2)/O(2) fit Blanc's law, in agreement with measurements. Calculations for He/N(2) and He/CO(2) are also consistent with observations and suggest why adding He to the working gas generally enhances FAIMS performance. As predicted, mixtures of gases with extremely disparate molecular masses and collision cross sections, such as He/SF(6), exhibit spectacular non-Blanc effects, which greatly improve the resolution and peak capacity of technique. Understanding FAIMS operation in gas mixtures is expected to enable the rational design of media for both targeted and global analyses.