The hydrophobic-polar (TIP) model is an abstract representation of the protein structure prediction problem, where hydrophobic interactions are assumed to be the major determinant of the folded state of proteins. This paper inquires into the constraint-handling design issue of metaheuristics, which is crucial when dealing with such a challenging and highly constrained combinatorial optimization problem. A new constraint-handling strategy for the TIP model, based on multiobjective optimization concepts, is here proposed. The multiobjective approach for handling constraints in this particular problem is explored for the first time in this study, to the authors’ knowledge. Using a basic genetic algorithm and a large set of test instances for the two-dimensional TIP model (based on the square lattice), the proposed multiobjective strategy was evaluated and compared with respect to commonly adopted techniques from the literature. On the one hand, through such a comparative analysis it was possible to demonstrate the suitability of the proposed multiobjective strategy. On the other hand, the results of this study provide further insight into whether infeasible protein conformations should be allowed or prevented during the metaheuristic search process, which has been a subject of concern in the specialized literature.