Many real-world optimization problems involve multiple objectives and constraints, often represented by discontinuous, non-differentiable or black-box functions. With multiple conflicting objectives, a global optimum does not exist. The optimization in these cases requires computing a set of non-dominated solutions, the Pareto set. Population-based techniques have proven effective for multiobjective optimizations, but often suffer from slower convergence in the presence of numerous constraints. Estimation of Distribution Algorithms (EDAs) are optimization techniques that update the population by sampling a probability density function fitted to the best individuals at each iteration. The Multiobjective Estimation of Distribution Algorithm based on a Gaussian Mixture Model (MOEDA-GMM) employs a Gaussian Mixture Model (GMM), fitted through the Expectation-Maximization (EM) algorithm, as its probability density function. This method shows results comparable to consolidated multiobjective optimization techniques, such as Matlab’s paretosearch and gamultiobj, as measured by the hypervolume metric. These findings highlight the potential of MOEDA-GMM as an effective tool for constrained multiobjective optimization.