Artificial Neural Networks and Random Forest are two machine learning models intensively used in medical field, providing accurate results in many circumstances that require predictions or decision-making. Nevertheless, it is not established yet which of these two algorithms is more suitable for a specific context. The work presented in this article investigates how machine learning methods can help in real medical applications. The main objectives of this research were to use, to evaluate and to compare artificial neural networks and random forest. Both models were trained and evaluated using the same dataset and identical performance metrics to ensure a fair and consistent comparison. This study aimed to highlight both the advantages and potential limitations of the two algorithms, while also exploring their applicability within a specific medical domain – namely, the early detection of diabetes. Both models proved remarkable results, and the comparative analysis conducted revealed that each algorithm is characterized by distinct strengths in specific aspects of the classification task.