A new methodology based on Henrys law is proposed for modeling the release of fragrances from a simplified matrix commonly used in consumer products formulations. For that purpose, different mixtures were formulated containing one, two, three, or four fragrance ingredients diluted in dipropylene glycol (simplified matrix). Headspace concentrations were measured to estimate Henrys constants (H) for each fragrance component in all mixtures. The individual Henrys constants for multicomponent fragrance mixtures were also predicted from the ones measured for each single compound diluted in the matrix. Furthermore, we used a model that combines the UNIFAC group-contribution method with the modified Raoults law and the psychophysicals Stevens power law and strongest component model to predict the perceived odor intensity and character, respectively. Results showed a strong linear relationship between experimental H for single fragrances and experimental H for binary (r(2) = 0.998), ternary (r(2) = 0.997), and quaternary (r(2) = 0.996) fragrance mixtures. This new approach can bring a relevant advantage to the preformulation process by reducing time and cost associated with trial-and-error experiments.