Preeclampsia, a hypertensive disorder complicating pregnancy, is a leading cause of maternal and fetal morbidity and mortality characterized by elevated blood pressure and multi-organ dysfunction affecting hepatic and renal functions. Early detection of preeclampsia is critical as the disorder can progress rapidly to life-threatening conditions such as eclampsia and stroke. However, traditional diagnostic methods, primarily based on clinical signs, often identify the condition too late in its progression. This study introduces an advanced machine learning (ML) approach to enhance early detection by analyzing a comprehensive dataset of maternal health indicators including maternal age, parity, blood pressure, and cardiovascular and metabolic biomarkers. The study examines several ML algorithms—Logistic Regression, Decision Trees, and Random Forest—employing hyperparameter tuning via Randomized Search CV to optimize performance. Data preprocessing involved managing missing data, addressing outliers, and applying feature engineering techniques such as creating new variables like blood pressure variability. Feature selection focused on retaining the most clinically relevant indicators, with models trained using cross-validation and an 80–20 training–testing split. The Random Forest model emerged as the most effective, outperforming Logistic Regression and Decision Trees, particularly in handling non-linear interactions between features. Evaluation metrics including accuracy, precision, recall, F1-score, ROC-AUC, confusion matrix, logarithmic loss, and Matthews Correlation Coefficient (MCC) were used to assess model performance. Notably, the Random Forest model achieved superior accuracy (84%), high F1-scores, and robust AUC values, highlighting its strong predictive power and reliability in differentiating high-risk from low-risk pregnancies. The inclusion of clinically relevant features such as blood pressure variability and maternal comorbidities significantly enhanced the model’s performance. Additionally, feature importance analysis provided insights into key predictors, reinforcing the model’s clinical utility. Ultimately, this ML-based tool shows great promise for integration into clinical practice, with the potential to improve maternal and fetal outcomes by enabling timely interventions. Future research will focus on incorporating real-time clinical data and conducting broader population validation to ensure the model’s generalizability.