The impact of assimilating wind data on tropical cyclone track forecasts is studied in a series of experiments using a nondivergent barotropic model. In particular, the impact of wind observations on track forecasts in significantly baroclinic environments is investigated using both four-dimensional variational assimilation (4D-VAR) and ensemble Kalman filter (EnKF) techniques. Results suggest that assimilating wind data can improve forecasts in situations where barotropic track models tend to perform the poorest.