The design and application of a nonlinear Generalized Likelihood Ratio (GLR) algorithm for target maneuver detection and estimation for short-range air-to-air missile scenarios is addressed. The problem, which is inherently nonlinear, is first reformulated into a linear problem by preprocessing the measurements. The maneuver detection algorithm then consists of a Kalman filter that is tuned to track the target under nonmaneuvering conditions and a GLR which monitors the innovations process of the filter to determine if a maneuver has occurred. Maneuver estimation is accomplished via maximum likelihood techniques and, once a maneuver is estimated, the states of the Kalman filter and their error covariances are suitably adjusted.