In this article, we propose a new technique for enabling multivariable volume fraction analysis by microwave resonator-based sensors. We address this issue using material dispersion and its nonlinear impact on the harmonic frequency shift. The concept is utilized to determine the volumetric fraction of multicomponent mixture selectively. Since the cutoff frequency of the permittivity spectrum is different in materials, by a proper selection of permittivity sampling frequencies, independent features are measured which consequently results in multivariable sensing capability. Debye model as a simplified presentation of the Cole-Cole relation plays a significant role in distinguishing the components in the mixture. For the proof of concept, mixtures of methanol, ethanol, and propanol are used as the material_under the test (MUT) and it is shown that the volume fraction of the components can be selectively distinguished. The concept in theory and experiment is verified and the formulation is explained in detail. The proposed concept paves the way toward the widespread use of microwave resonant sensors for selective real-time monitoring of chemical mixtures.