Observing System Simulation experiments were performed to examine whether data from above 45,000 ft could potentially improve hurricane track forecasts. The hurricane prediction system of the National Centers for Environmental Prediction was used to generate global ensemble forecasts which include hurricane track forecasts made using the Geophysical Fluid Dynamics Laboratory hurricane model embedded in the global model. One ensemble forecast was labeled the "truth" while another ensemble member was used as the first guess. Data from the truth run was assimilated into the first guess and the forecast was then run out to 72 hours. Data were assimilated from within 3000 km of the center of the "true" hurricane using three different strategies: assimilating data below 45,000 ft only; assimilating data between 45,000 and 60,000 ft., and assimilating all data up to 60,000 ft. A total of 11 cases were examined. In all cases, data from below 45,000 ft. resulted in large improvements in track forecasts, while data above 45,000 ft. resulted in little or no improvement. This suggests that observations of hurricanes at and below the operating ceiling of the NOAA G-IV hurricane reconnaissance aircraft suffice for optimal hurricane track prediction.