The accurate age estimation of human is crucial in several cases and fields. Bone Age Assessment (BAA) can be an effective method for human age estimation for live and dead human. Several manual methods have been proposed to achieve this task; however, they are time consuming and error-prone. In this paper, an automated bone age assessment system is proposed based on the concept of transfer learning for feature extraction. In the proposed system, the hand X-ray images are preprocessed, and the discriminant features are extracted using a pre-trained, fine-tuned deep neural networks (AlexNet and ResNet-101). Finally, the age group of the hand X-ray image is determined using a number of classification models including decision tree, k-nearest neighbour, linear discriminant, and support vector machine. The proposed system is assessed using the RSNA Bone Age dataset. The obtained results have shown that the ResNet-101-based features are more effective than the corresponding AlexNet-based features. In addition, the decision tree classifier is better than the remaining classifiers with classification accuracy up to 100%.