Galvanostatic intermittent titration experiments were performed in three-electrode cells to characterize the effect of C/2, 2-C and 4-C charge rates on the observed lithium diffusion coefficient. As part of the data analysis process, we compared the classic Weppner-Huggins analysis of polarization data with a newer (Wang et al.) analysis method for depolarization data. At low values of x in LixC6, both analysis methods showed the same general trend in the apparent lithium diffusion coefficient, 4-C > 2-C > C/2. The two techniques differed in the magnitude of the estimated diffusion coefficient by about a factor of 100. The observed increase in diffusion coefficient does not last over a large compositional range. Since the estimates from the method of Weppner and Huggins may contain artifacts due to the use of particulate electrodes and high charge rates, the method of Wang et al. may produce better values.