Different criteria exist for the classification of the metaheuristics. One important classification is: improvement metaheuristics and constructive. On the one hand improvement metaheuristics, begins with an initial solution and iteratively improves the quality of the solution using neighborhood search. On the other hand, constructive metaheuristics, are those in which a solution is built from the beginning, finding in each iteration a local optimum. In this article, we to compare two constructive metaheuristics, Ant Colony Optimization and Intelligent Water Drops, by solving a classical NP-hard problem, such like the Set Covering Problem, which has many practical applications, including line balancing production, service installation and crew scheduling in railway, among others. The results reveal that Ant Colony Optimization has a better behavior than Intelligent Water Drops in relation to the problem considered.