Brain tumor segmentation is one of the challenging image segmentation problem. Among the various architectures of CNN used for brain tumor segmentation from MRI images, we compare multi-view 2D CNN and multi-view 3D anisotropic CNN on the popular BraTS dataset. We computed four metrics to measure their performance on 10 test data. The first approach showed a mean sensitivity of 0.816 whereas second approach outperforms with mean sensitivity of 0.9217. Other metrics for both approaches achieved comparable results. Although both approaches consider all orthogonal planes, anisotropic CNN takes into account both global and local features efficiently thereby giving better results.