Chromosome classification is critical for karyotyping in abnormality diagnosis. To expedite diagnosis process, we present a novel method named Varifocal-Net for simultaneous classification of chromosome's type and polarity using deep convolutional networks. The approach consists of one global-scale network (G-Net) and one local-scale network (L-Net). It follows two stages. The first stage is to learn both global and local features. We extract global features and detect finer local regions via the G-Net. With the proposed varifocal mechanism, we zoom into local parts and extract local features via the L-Net. Residual learning and multi-task learning strategies are utilized to promote high-level feature extraction. The detection of discriminative local parts is fulfilled by a localization subnet of the G-Net, whose training process involves both supervised and weekly-supervised learning. The second stage is to build two multi-layer perceptron classifiers that exploit features of both two scales to boost classification performance. Evaluation results from 1909 karyotyping cases demonstrate that our Varifocal-Net achieved the highest accuracy of 0.9805, 0.9909 and average F1-score of 0.9771, 0.9909 for the type and polarity task, respectively. It outperformed state-of-the-art methods, demonstrating the effectiveness of our Varifocal mechanism and multi-scale feature ensemble.