Brain tumor detection employs sophisticated deep learning methods to construct a robust system that can classify MRI images into two categories: those depicting tumors and those without. Through the use of convolutional neural networks (CNNs), the project endeavors to support radiologists by offering an automated tool for early and precise tumor identification. The approach involves thorough pre-processing of images, including cropping, resizing, and normalization, followed by data augmentation to improve model training on a limited dataset. The system has undergone training and validation on a carefully curated dataset, achieving a notable accuracy of 91.7% and other vital performance measures like precision, recall, and F1 score.