In today's competitive retail market, dynamic pricing has become a crucial component of successful business strategies. This paper explores how predictive analytics and machine learning can be effectively utilized to optimize retail prices and maximize profits. Traditional static pricing methods fail to adapt to the fast-changing market conditions influenced by seasonality, competition, and consumer behavior. The study implements and compares three supervised machine learning models—Linear Regression, Random Forest, and Support Vector Regression (SVR)—to predict optimal pricing using real retail data. After extensive training, tuning, and validation, the Random Forest model emerged as the most accurate, achieving an R² score of 0.9897, indicating a near-perfect fit. This research highlights how data-driven approaches enhance decision-making accuracy and pave the way for real-time dynamic pricing in modern retail environments.