This paper describes a variant of quantum-behaved particle swarm optimization (QPSO) algorithm named C-QPSO for solving global optimization problems. In C-QPSO the QPSO is revised by including a novel crossover operator to enhance the algorithm's ability to escape from local optima. The experimental results on test functions demonstrate that the proposed hybrid optimization algorithm performs much better than PSO, original QPSO and two QPSO variants in terms of their convergence and stability.