Brazil is a country of continental dimensions and one of the most populous in the world. With companies operating in a wide range of economic sectors, the country has many accidents and occupational diseases annually. Therefore, it is of fundamental importance to understand the profile of occupational accidents in the country and, using machine learning techniques, to predict possible occurrences of accidents. In this article, we conduct an extensive exploratory analysis of occupational accidents in Brazil. We then propose a systematic extraction and transformation of variables from several statistical databases to obtain a single table, which is used in machine learning model training. Finally, we propose constructing regression models - linear regression, support vector machine (SVM), XGBoost, and LightGBM - to predict occupational accidents in the Brazilian states. In this work, the predictions of gradient boosting algorithms have higher \(R^2\) and lower errors.