We present a technique for determining parameters of any activity coefficient model from the times at which resonances appear in light scattering spectra from an evaporating multicomponent droplet of known initial composition. Using individual component droplet evaporation data along with assumed values for the parameters of a chosen activity coefficient model, the size and the composition of the droplet as functions of time are theoretically predicted, and the times at which resonances appear are calculated from Mie theory. By minimizing the errors between the calculated and observed resonance appearance times, the optimum parameter values of the activity coefficient model are established. The technique has been applied to examine binary and ternary systems consisting of diethyl phthalate, dipropyl phthalate, and dimethyl phthalate. In addition, the technique has been validated by comparing the prediction from the optimum parameter values with independently determined activity coefficient data for binary mixtures of dioctyl phthalate and dimethyl phthalate.