Maintenance on transformers is currently carried out on a monthly basis, which activities are carried out manually so that it is difficult to detect transformers experiencing abnormal conditions if there is a disturbance outside the maintenance schedule. Therefore, this research was carried out with the aim of making a prototype of a transformer condition monitoring system and monitoring parameters on the transformer remotely through the Blynk application in real time. In making this tool, the PZEM-004T sensor is used to read electrical parameters, the DS18B20 sensor to read the temperature on the transformer winding and the Arduino MEGA 2560 With WiFi Built-in - ESP8266 to process the data read by the sensor which is then sent to the Blynk application. The results of this study indicate that the average error value for the comparison of current readings is 0.68%, the average error value for the comparison between phase to neutral voltage readings is 0.91%, the average error value for the comparison of voltage readings between phases is 1, 00%, the average error value of the Cosφ reading comparison is 0.02% and the average temperature reading comparison error value is 1.73%.