Abstract: A quantum wolf pack algorithm is proposed to solve 0-1knapsack problems. Defining probability positions and accurate positions of particles in population refer to quantum encoding. Quantum rotation is used to orient probability positions of wolves to the global optimal position, and then completing the mapping from probability position to accurate position by quantum collapse, these operators reconcile the orientation and randomness. There are 8 classical and 3 high-dimensional knapsack problems employed to test the proposed algorithm, and then compared the performance of this algorithm with other algorithm. The statistical results demonstrate the effectiveness and global search capability on knapsack problems especially for high level cases.