This research investigates the development of a machine learning-based model for crop production and fertilizer recommendations aimed at enhancing agricultural productivity in North Mecha Woreda, Ethiopia. Recognizing agriculture's critical role in the Ethiopian economy, the study addresses challenges faced by farmers, such as the need for accurate crop yield predictions and effective fertilizer use. The methodology involves several interdependent phases: data collection encompasses a dataset of 2,654 records, including meteorological, soil, and historical crop yield data; data preprocessing ensures the quality and consistency of this dataset; model selection evaluates various algorithms, including Support Vector Machine (SVM), Multiple Linear Regression (MLR), and Artificial Neural Network (ANN); model development focuses on training these selected algorithms; and model evaluation employs performance metrics such as Mean Absolute Error (MAE), Mean Square Error (MSE), and R² (Coefficient of Determination). The interdependency among these phases is critical; the accuracy of crop type prediction directly influences yield estimates and informs appropriate fertilizer recommendations. A precise understanding of which crops are being cultivated allows for tailored fertilizer usage, optimizing both yield potential and resource use. Multi-output regression and classification approaches are utilized for crop type prediction, while Support Vector Regression is applied for fertilizer recommendations, and Artificial Neural Networks are employed for crop yield prediction. Experimental results show that the ANN achieved an R² of 0.975 for crop yield prediction, while the SVM excelled in crop type and fertilizer recommendations, recording an MAE of 0.0616 and MSE of 0.00566. Keywords: Crop Production, Fertilizer Recommendation, Machine Learning, Support Vector Machine, Artificial Neural Network, Yield Prediction