Palm oil is an agricultural commodity that has an important role in the global economy. Palm oil is obtained from the fruit of the oil palm tree (Elaeis guineensis) which grows in tropical regions, especially in countries such as Indonesia, Malaysia, Thailand and several West African countries. Palm oil has a variety of uses in the food, cosmetics and fuel industries, making it one of the most traded commodities in the world. Palm oil price fluctuations have a significant influence on the economy in producing and consuming countries. Therefore, a time series analysis is needed that can predict fluctuations caused by certain conditions. This analysis is carrying out nonlinear analysis using the SETAR (Self Exciting Threshold Autoregressive) method on palm oil prices to obtain a prediction model and prediction results for palm oil prices. The SETAR model is a special case part of the Threshold Autoregressive (TAR) model. The SETAR model threshold is a lag value of the series itself or the endogenous variable. Analysis carried out using the SETAR method produces a SETAR (3,1,1) model with threshold (r) = 0.01626070 where the fit value approaches the actual data value and the predicted value follows the actual data pattern