A new algorithm for pass adaptation in plate rolling is developed to improve thickness accuracy of plate products. The feature of the algorithm is that it uses the measured data rather than the schedule calculated data in adaptation, which leads to notable improvement in prediction accuracy of the rolling parameters and thickness accuracy of products can be improved accordingly. Results show that this adaptive algorithm is effective in practice. A new algorithm for pass adaptation in plate rolling is developed to improve thickness accuracy of plate products. The feature of the algorithm is that that it uses the measured data rather than the schedule calculated data in adaptation, which leads to not able improvement in prediction accuracy of the rolling parameters and thickness accuracy of products can be improved accordingly. Results show that this adaptive algorithm is effective in practice.