This article first presents an overview of the recent advances in the analysis and prediction of tropical cyclones through assimilating reconnaissance aircraft observations. Many of these advances have now been implemented in operational and experimental real-time hurricane prediction models. These advances are made possible through improved methodologies including more efficient quality control and data thinning, advanced data assimilation techniques that use ensembles to estimate flow-dependent error covariances, and improved numerical models running at convection-permitting resolutions, along with the availability of massively parallel computing.
 Impacts of aircraft reconnaissance observations on hurricane prediction are then exemplified using a continuously cycling regional-scale convection-permitting analysis and forecast system based on the Weather Research and Forecasting (WRF) model and the ensemble Kalman filter (EnKF). In comparison to the non-reconnaissance experiment that assimilates only conventional observations, as well as to the WRF forecasts directly initialized with the global operational analysis, the cycling WRF-EnKF system with assimilation of aircraft flight-level and dropsonde observations can considerably reduce both the mean position and intensity forecast errors for lead times from day 1 to day 5 averaged over a large number of forecast samples including the real-time implementation during the 2013 Atlantic hurricane season. These findings reaffirm the added value and need for maintaining and maybe expanding routine airborne reconnaissance missions for better tropical cyclone monitoring and prediction.