Precise prescription drug selection tailored to patient profiles is vital for ensuring safety, efficacy, and individualized therapy, ultimately enhancing clinical results and patient wellbeing. Traditional manual approaches are inefficient, slow, and susceptible to errors, making them impractical for high-volume healthcare environments. This research examines machine learning techniques to forecast ideal medications from patient features such as age, blood pressure, gender, cholesterol levels, medication history and Na/K ratio. We assessed six classifiers-including ensemble methods, linear models, kernel-based classifiers, and probabilistic models-using metrics like- precision, recall, accuracy, and F1-score. Linear and kernel classifiers topped the list at 85% accuracy, showing strong balance across drug categories. These outcomes demonstrate AI's promise for streamlining clinical choices, cutting manual effort, speeding up prescriptions, curbing errors, and boosting care quality. Looking ahead, integrating bigger datasets, neural architectures, and interpretable AI could further improve reliability and real-world adoption. Overall, this work advances data-driven personalization in medicine.