Cognitive modeling is an emergent field in psychology that allows to model human behavior with the help of mathematical models. The Drift Diffusion Model (DDM) represents the decision behavior of human beings in two-choice reaction-time tasks. The goal in cognitive modeling is to fit a model to data to describe the processes found in the data as close as possible. In order to fit the DDM to data, several estimation methods and software have been introduced. One method to quantify and estimate parameters of the DDM is the maximum likelihood approach. Particle Swarm Optimization (PSO) is a direct optimization algorithm which can find minima of parameters within a large search space. As such, it is well suited to find parameters of the DDM that minimize the negative log-likelihood. PSO has rarely been used so far in fitting the DDM but this paper shows its well-suited application with replicable, easy understandable pseudo-code. Adaptations of the PSO algorithm which introduced a rebounding measure as well as an anarchy measure highly increased its efficiency. As such, the paper offers a novel application of PSO in the realm of cognitive modeling.