This thesis deals with the mathematical modeling, the estimation, and the control of the contour evolution in heavy-plate rolling. Reversing mill stands are used to reduce the thickness of heavy plates in consecutive rolling passes. During the rolling process, asymmetric rolling conditions in the lateral direction may lead to a deviation between the actual and the desired plate contour. This may cause a reduced product quality, product rejects, and in the worst case even damaged plant components around the rolling mill. Asymmetric rolling conditions are generally unknown and cannot be compensated in advance to prevent the plate from cambering. Therefore, a useful approach is to apply feedback if a shape defect occurs. Clearly, this requires a measurement of the contour of the plate. Precise measurements of the contour (longitudinal boundaries and shape of the head and tail end) can be used to optimize the adjustment of the mill stand to reduce the camber. First, the estimation of the plate contour is discussed. Infrared cameras mounted at the ceiling of the rolling mill are used to capture images of the plate during the rolling passes. A threshold-based edge detection is performed in the infrared bitmaps. The detected edges are then used in an optimizationbased approach which utilizes the restrictions of the movement of the plate being clamped in the rolling gap. Herein, a polynomial representation of the longitudinal boundaries of the plate is estimated based on the detected edges of several consecutive images. In addition to the contour of the plate, the translational and angular movement of the plate is estimated by the presented approach. A validation by means of a downstream contour measurement device shows a high accuracy of the proposed contour estimation method. Second, a static model of the contour evolution is derived. The continuummechanics-based model predicts the contour of the plate after the rolling pass based on the contour before the rolling pass and the input and output thickness profiles. It is used to analyze the effects of temperature gradients in the lateral and during the standard production process are used to validate the model. Moreover, a model covering the relation between the contour and the movement of the plate is presented. In particular, the model links the curvature and the angular velocity at the entry of the rolling gap with the curvature and the angular velocity at the exit of the rolling gap. The model is validated using the angular velocity and curvature of the plate obtained from the contour estimation approach. The mathematical models are used in different control approaches for the reduction of contour errors. First, a feedforward control strategy to determine the required asymmetry of the rolling gap height to compensate contour errors in single passes is discussed. The optimization-based approach utilizes the continuummechanics-based model of the contour evolution. The control objectives can be changed in an intuitive manner by changing weighting factors of the objective function and input constraints are systematically incorporated. Moreover, the asymmetric compliance of the mill stand as a function of the rolling forces is identified and compensated based on the desired rolling forces. The resulting asymmetry of the output thickness after each pass is estimated and used to compensate disturbances affecting the asymmetry of the rolling gap. Furthermore, feedback control during the rolling pass is discussed. The measurement of the contour of the plate is subject to a transport delay. Hence, the presented approach utilizes the delay-free measurement of the angular movement of the plate in a two degrees-of-freedom control structure containing a Smith-predictor. Furthermore, a proof of the robust stability of the proposed control concept is presented. In general, compensating a contour error results in an inhomogeneous thickness profile in the lateral direction. Therefore, a method to determine a rolling schedule covering several rolling passes is presented which achieves both the desired contour and a homogeneous thickness profile of the final product. Finally, simulation results and measurements for the proposed control approaches are shown. The influence of weighting and tuning factors on the control behavior is discussed by means of simulations. Measurements from the considered industrial rolling mill show that the proposed measures can significantly improve the contour of the rolled plates.