This paper explores the method of reassignment for extracting instantaneous frequency attributes from seismic data. The reassignment method was first applied to the spectrogram by Kodera, Gendrin and de Villedary and later generalized to any bilinear time-frequency or time-scale representation by Auger and Flandrin. Key to the method is a nonlinear convolution where the value of the convolution is not placed at the center of the convolution kernel but rather reassigned to the center of mass of the function within the kernel. The resulting reassigned representation yields significantly improved component localization. In this paper we will study the impact of the reassigned time-frequency representation on our ability to reliably estimate instantaneous frequency for a given seismic signal.