Preeclampsia complicates more than 7% of pregnancies worldwide and poses risk for the health of both mother and fetus. Because preeclampsia can be life-threatening and early in its course, has mild symptoms ranging widely both from woman to woman and even within the same patient over time, patients with any signs or symptoms of preeclampsia should also have close follow-ups. If any alarming symptoms are noted during pregnancy, one should consult healthcare providers on an urgent basis. The assessment and monitoring needed for managing preeclampsia can place a considerable financial strain on healthcare systems. In light of this, the research utilized machine learning techniques, specifically the XGBoost and Particle Swarm Optimization (PSO) algorithms, to enable early detection of preeclampsia. The study involved 168 cases of pregnant women with preeclampsia and 336 control samples, using 14 features. The results indicated that the Hybrid PSO with XGBoost Algorithm enhanced the accuracy of the XGBoost model. Initially, XGBoost achieved an accuracy of 95.46%, with a 70% training and 30% testing data split. After applying the Hybrid PSO, accuracy improved to 97.46% when the population size was 10, the maximum number of generations was 40, and the inertia weight was 1.0, showing a 2% increase. Notably, seven features had a significant impact, with correlation weights of 10% or higher. These findings are significant for the healthcare sector, particularly in preventing and managing pregnancy-related complications, and can support technology-based early detection of preeclampsia.