This paper presents an evolutionary algorithm that incorporates a multilevel pairing strategy to solve single and multiobjective optimization problems. The algorithm is based on nondominance of solutions separately in the objective and the constraint space and uses cooperative mating strategies between solutions. Since the methodology is based on nondominance separately in the objective and the constraint space, scaling and aggregation affecting conventional penalty function methods for constraint handling does not arise. The proposed cooperative and intelligent pairing strategies result in mating between solutions that are good in objectives with those that are good in constraint satisfaction, thus helping to speed up convergence. The diversification mechanism in the algorithm is based on niching that results in a wide spread of solutions in the parametric space. Three constrained multiobjective design examples and a single objective optimization problem with continuous and mixed variables are used to illustrate the performance of the proposed algorithm.