In this study, a methodology is presented for segmenting brain tumor using the U-Net deep learning approach. The accurate segmentation of brain tumors is critical for their diagnosis and treatment, and the U-Net architecture has demonstrated promising results in biomedical image segmentation tasks. The proposed methodology consists of three primary stages: Data Preparation, Data Pre-processing, and Brain Tumor Segmentation. The model is trained on a BraTS 2020 dataset of MRI brain scans that includes both Tumor and Non- Tumor samples. The U-Net architecture is utilized, and the achieved results shows a maximum accuracy of 99.6% for various epochs. The proposed approach has substantial potential for improving the accuracy of brain tumor diagnosis and treatment, leading to better patient outcomes. This research highlights the effectiveness of deep learning in medical image segmentation tasks and underscores the importance of accurate segmentation in brain tumor diagnosis and treatment.