Flooding is one of the disasters that often occurs in Indonesia, especially in areas with high rainfall, such as Malang City. To improve the effectiveness of early warning, this research develops an Early Warning System (EWS) information dissemination system based on the Internet of Things (IoT). This system combines Variable Message Sign (VMS) as a visual display media and a Telegram bot to provide real-time notifications to the public. NodeMCU ESP8266 is the main microcontroller connected to a WiFi network to access data on the server. Tests were conducted to assess the accuracy of VMS synchronization, the speed of notification response, and the effectiveness of the buzzer in providing warnings. The test results show that synchronizing two VMS with different internet networks has an average time difference of 1.16 seconds, while with the same network, it is only 0.03 seconds. In addition, the VMS and Telegram bot can deliver information quickly and accurately, while the buzzer functions according to the set warning level. With this system, the dissemination of flood early warning information becomes more effective and easily accessible to the public, especially those in vulnerable areas or on the move.