Soil moisture related to water content which a factor that affects the plant growth. The process of watering plants is generally done manually regardless of the volume of water needed by plants. This re research discussed about an automated prototype and a system that have the function of watering plants based on the soil moisture level. The method used is prototyping which is suitable with the research purpose. The prototype and systems built with microcontroller, soil moisture sensors, relay and solenoid valve, which integrated with the IoT platform Blynk apps and Thingspeak. The process starts from the detection of soil moisture by the sensor. If soil moisture value is detected on 30% - 35%, then the device activates the watering function by opening the valve from the solenoid valve to drain water to the pipe. When the soil moisture detected more then 35% , the device stops the watering function. ThingSpeak IoT platform, used to display moisture percentage data in graphical form. Blynk apps provide notification features to the user's smartphone when the watering device is activated or deactivated. Based on the test scenario performed, it was found that the percentage of soil moisture with an initial value of 30% - 35% increased to 68.2%, after the watering process. Each component of the device and system has been tested and functioning according to the purpose, so the system has the potential to be used in the process of watering the plants automatically.Keywords: automatic watering system, microcontroller, soil moisture, sensor, IoT.