Skin diseases are conditions that affect the skin, including infections, inflammatory conditions, autoimmune disorders, and skin cancers. They can be diagnosed by a dermatologist or a healthcare professional through a physical examination, medical history, and, in some cases, laboratory tests or skin biopsies. This skin disorders may be detected using deep learning algorithms by training on large datasets of healthcare images to learn patterns and features that are indicative of different skin diseases. EfficientNet-B6 and EfficientNet-B7 deep learning models are used to identify different skin diseases like eczema, benign keratosis-like lesions, basal cell carcinoma and melanoma. The proposed neural network model shows high performance on image classification tasks. To identify skin diseases using the different EfficientNet models, the medical images of skin are fed into the model as input. The models extract features and patterns that are indicative of a particular skin disease, and the resulting characteristics are then utilized to categorize the image as belonging to a particular skin disease class. With the right training and testing, the proposed models can provide accurate and efficient identification of skin diseases, potentially aiding in early detection and treatment. By evaluating the performance of all these models using metrics like Precision, Recall, F1-score and accuracy. EfficientNet-B7 achieved an accuracy of 97%.