Lemon trees are an essential component of global agriculture, providing significant economic and health benefits. Lemons are widely used across various industries, including food production, pharmaceuticals, and cosmetics, making them a valuable crop not only in Egypt but worldwide. However, lemon plants are highly susceptible to a variety of diseases, which can significantly impact their yield and quality. In this research, a deep learning-based approach is applied to diagnose lemon leaf diseases using a dataset of 1354 images categorized into nine different disease classes. To enhance the dataset and improve model generalization, various preprocessing techniques and data augmentation methods were employed, expanding the dataset to 8000 images. This enriched dataset was then used to train and evaluate several recent deep learning models, including Convolutional Neural Networks (CNN), ResNet50, VGG19, EfficientNetB0, and MobileNet. These models were assessed based on their classification accuracy, robustness, and computational efficiency. The experimental results demonstrated that after augmentation, the models achieved outstanding classification performance, with the highest accuracy reaching 99.8% and an F1-score of 1.0. These findings highlight the effectiveness of deep learning techniques in accurately identifying lemon leaf diseases, which can contribute to early detection and better management strategies. By leveraging advanced AI-driven solutions, this research provides a scalable and automated approach to assist farmers and agricultural experts in disease diagnosis, ultimately improving crop health, productivity, and sustainability.