Background: Modern agriculture operates within an increasingly unpredictable environment, influenced by dynamic market fluctuations and environmental variability. Timely and accurate prediction of crop prices is essential to support data-driven decision-making for farmers, policymakers, and stakeholders in the agricultural supply chain. Objective: This study presents a comprehensive machine learning framework aimed at forecasting crop prices by integrating environmental, economic, and logistical variables. The primary objective is to enhance agricultural profitability and sustainability through precise, data-informed insights. Methods: A diverse dataset was compiled, encompassing features such as temperature, precipitation, supply and demand metrics, transportation costs, fertilizer application, pest infestation levels, and market competitiveness. Advanced feature engineering techniques were applied to preprocess and refine the input data. Several machine learning models, including Linear Regression, AdaBoost, Support Vector Machines, Random Forests, and XGBoost, were developed and evaluated for their predictive accuracy. Results: Among the evaluated models, XGBoost outperformed the others by delivering the highest accuracy in price forecasting. Its capability to model complex, non-linear relationships and capture intricate feature interactions proved critical for reliable predictions. The enhanced precision offered by XGBoost enables stakeholders to make informed decisions, contributing to increased profitability and optimized resource allocation. Conclusions: The proposed XGBoost-based crop price prediction framework demonstrates robust performance in real-time agricultural forecasting scenarios. By incorporating a wide range of environmental and market variables, the model significantly reduces uncertainty in the agri-value chain, thereby supporting sustainable farming practices and improving economic resilience.