A new technique is presented for compensating radome-induced line-of-sight (LOS) aberrations in active radar-guided homing missiles based on a novel nonlinear estimation algorithm. Decoupling of radome errors from the nominal missile state is achieved by implementing a two-step estimator that utilizes non-linear measurements of the radomecorrupted azimuth and elevation LOS angles. Elaborate aerodynamic and kinematic models are used to establish a realistic interception setup, which constitutes a basis for a thorough performance evaluation and comparison of the new algorithm to a traditional extended Kalman filter. It is shown that the new approach requires neither specialized observability maneuver nor exogenous dithering to estimate the radome slopes. An unbiased, efficient estimates of radar slopes yielded by the new method permits considerable reduction of miss distance.