•A novel hybrid 3D deep residual convolutional U-Net (dResU-Net) architecture is proposed for brain tumor segmentation.•Low-level features are being preserved with high-level features for better segmentation within the encoder part.•The proposed model is cross-validated on an external cohort dataset to evaluate the robustness and generalizability of the architecture.•The proposed model used 3D multi-modal MRI for the automatic segmentation of brain tumor sub-regions.•The proposed study used a combined loss function based on Dice and Focal loss. Glioma is the most prevalentand dangerous type of brain tumor which can be life-threatening when its grade is high. The early detection of these tumors can improve and save the life of the patients. The automatic segmentation of brain tumor from magnetic resonance imaging (MRI) plays a vital role in treatment planning and timely diagnosis. Automatic segmentation is a challenging task due to the massive amount of information provided by MRI and the variation in the location and size of the tumor. Therefore, a reliable and authentic method to segment the tumorous region from healthy tissues accurately is an open challenge in the field of deep learning-based medical image analysis. This research paper presents an end-to-end framework for automatic 3D Brain Tumor Segmentation (BTS). The proposed model is a hybrid of the deep residual network and U-Net model (dResU-Net). The residual network is used as an encoder in the proposed architecture with the decoder of the U-Net model to handle the issue of vanishing gradient. The proposed model is designed to take advantage from low-level and high-level features simultaneously for making the prediction. In addition, shortcut connections are employed between residual network to preserve low-level features at each level. Furthermore, skip connections between residual and convolutional blocks in the proposed architecture are used to accelerate the training process. The proposed architecture achieved promising results with the average dice score for the tumor core (TC), whole tumor (WT), and enhancing tumor (ET) on the BraTS 2020 dataset of 0.8357, 0.8660, and 0.8004, respectively. To demonstrate the robustness of the proposed model in real-world clinical settings, validation of the trained model on an external cohort is performed on randomly selected 50 patients of the BraTS 2021 benchmark dataset. The achieved dice scores on the external cohort are 0.8400, 0.8601, and 0.8221 for TC, WT, and ET, respectively. The comparison of results of the proposed approach with the state-of-the-art techniques indicates that dResU-Net can significantly improve the segmentation performance of brain tumor sub-regions.