This study presents the application of various clustering techniques for the analysis of a traffic accident dataset. Methods such as Gaussian Mixture Model, Agglomerative Clustering, MiniBatchK-Means and K-Means are used to identify hidden patterns in the data, including features such as geographic coordinates, accident severity, cause and type of accident. The dataset is preprocessed by removing null values, encoding categorical variables, and robust scaling of features. Furthermore, PCA is applied to reduce the dimensionality of the dataset. The performance of the clustering techniques is evaluated using metrics such as the Silhouette Score, Davies-Bouldin Score and Calinski-Harabasz Score. The results indicate that K-Means, with 5 clusters, provides the best overall performance, according to the Elbow method and the evaluated metrics. Visualizations of 2D is included for a better interpretation of the clusters, highlighting the distribution and features of the groups formed. The code and dataset are available at Kaggle: https://www.kaggle.com/code/hvelesaca/smarttech-paper-id-42 , facilitating further research.