Transformers are in the generation, transmission, and distribution systems. In the transmission line, transformers are useful for lowering the voltage of 150 kV-20 kV. Viewed from an expensive economy, transformer performance needs to be maintained. Transformers will experience increased temperature due to loading which can cause a decrease in performance. At PT. Peony, which has 3 power transformers in one substation, the remaining life of the transformer has never been predicted, therefore the transformer needs to be predicted for its remaining life to optimize maintenance strategies to ensure the transformer remains in good condition. The method used is Long ShortTerm Memory that uses time series data to process non-linear data. The predicted remaining life of transformer 1 using the LSTM method in 2025 is 18.813 equivalent to 18 years 9 months 23 days, for transformer 2 it is 19.135 equivalent to 19 years 1 month 19 days, and for transformer 3 the remaining life is 19.380 equivalent to 19 years 4 months 17 days. Then transformer 1 produces an RMSE error of 0.561 and a MAPE of 2.856%, for transformer 2 RMSE value is 0.549 and MAPE is 2.888%, on transformer 3, RMSE is 0.934 and MAPE is 4.844%, with error below 10% shows that long short-term memory method is accurate when used to predict the transformer’s remaining life. Then the factors that affect remaining life of the transformer include loading, power usage, winding temperature (HV, LV, TV), and ambient temperature.