Accurately predicting obstetric outcome is a critical challenge in maternal-fetal medicine, with significant implications for improving healthcare practices and minimizing adverse events. A dataset comprising fetoplacental biometric features was used, with the response variable stratified into three distinct categories: newborn, intrauterine fetal death, and neonatal death. In this context, the present paper aimed to develop an application capable of predicting one of the three possible obstetric outcomes. Multinomial logistic regression models are commonly used to identify the most influential predictors, as they enable the modeling of relationships between independent variables and a categorical dependent variable with more than two categories. In that regard, within the scope of model determination, various variable selection techniques were implemented, with the Elastic net method, using an alpha of 0.1, showing the best performance. For the intrauterine fetal death category, the model identified four significant variables, achieving an AUC-ROC of 0.841, with female fetal gender emerging as the most contributive factor. Regarding the neonatal death category, seven relevant variables were selected, with the model yielding an AUC-ROC of 0.620. The most impactful variables in this category were Diameter2, male gender, and maternal age. Following this stage, the chosen model, integrating the multinomial logistic regression algorithm with the Elastic net technique, was employed in the development of a web application, which is available at https://obstetricoutcome.shinyapps.io/mo_uminho/ . This platform can be a significant contribution to obstetric practice, providing an intuitive tool for healthcare professionals.