According to data from the National Disaster Management Agency (BNPB), 629 landslides occurred in 2022, resulting in 318 fatalities, 459 displaced individuals, and extensive damage to 892 buildings and public facilities. To mitigate the impacts of such events, an early warning system for landslides based on Long Range Wide Area Network (LoRaWAN) was developed, enabling more effective monitoring and response in high-risk areas. This system integrates LoRaWAN technology with a suite of sensors, including a soil moisture sensor to track moisture levels, a Global Position System (GPS) sensor to provide location data, and an accelerometer to detect tilt and acceleration changes. Sensor data were transmitted to a gateway and monitored in real time via the Blynk application. Furthermore, the relationship between Spreading Factor (SF) values, transmission distance, Time on Air (ToA), and Packet Delivery Ratio (PDR) was examined to optimize system performance. The results indicate that SF 12 provides the most reliable performance in the context of early landslide detection. Data transmission in both emergency and scheduled modes was successfully achieved, with seamless integration of the gateway and Blynk platform. This research presents a robust framework for improving disaster mitigation efforts through early detection and monitoring systems.