E-cigarette devices are wide ranging, leading to significant differences in levels of toxic carbonyls in their respective aerosols. Power can be a useful method in predicting relative toxin concentrations within the same device, but does not correlate well to inter-device levels. Herein, we have developed a simple mathematical model utilizing parameters of an e-cigarette’s coil and wick in order to predict relative levels of e-liquid solvent degradation. Model 1, which is coil length/(wick surface area*wraps), performed in the moderate-to-substantial range as a predictive tool (R2 = 0.69). Twelve devices, spanning a range of coil and wick styles, were analyzed. Model 1 was evaluated against twelve alternative models and displayed the best predictability. Relationships that included power settings displayed weak predictability, validating that power levels cannot be reliably compared between devices due to differing wicking and coil components and heat transfer efficiencies