In general, sustainable agriculture is increasingly challenged by the scarcity of water resources and recurrent plant infections, yet most of the existing systems address irrigation and detection separately. Currently, most smart irrigation systems control only water consumption, and, as a rule, plant disease detection operates independently. Most small farmers cannot afford or access such an integrated, user-friendly system. This paper describes an IoT-enabled Smart Plant Health Monitoring System that integrates automated irrigation with real-time detection of plant diseases and pests using machine learning. Watering is controlled by the proposed system, taking into account the amount of water required by the plants due to soil moisture and environmental sensors. An ESP32-CAM captures images of plants; through image processing, a lightweight CNN identifies diseases and pests. The whole system is integrated by a dashboard developed on Streamlit, where farmers can observe, receive suggestions, and act if necessary. Tests show that the system works well at recognizing diseases and pests, and it uses water more efficiently than traditional methods. By combining watering and plant health monitoring into one low-cost system, this solution is especially helpful for small and medium farms. It helps save water, increase crop production, and support farming in a more sustainable way.