Abstract: Based on the genetic algorithm with embedded fuzzy rules and a 0-1 integer planning model thus set up, an online optimization algorithm is developed for interstand tension setting during tandem cold rolling. The algorithm can improve productivity and products' quality with energy saved in addition to its high computing speed and online self-learning feature. The result of interstand tension setting optimization on tandem cold rolling mill for 1767 mm strip indicated that the genetic algorithm with embedded fuzzy rules can optimize online the interstand tension setting decision and play an reference role in exploring further the high-efficiency running of the tandem mill.