Abstract
Cassava are one of the staple foods in place of rice. However, today almost no one consumes cassava as a staple food substitute for rice. Indonesia is ranked third as the world's largest producer of cassava. Therefore, to maintain the value of yam production, the author will look for linear regression models as well as the best models with factors that are harvest area and productivity. Productivity is defined as the result of a comparison between the area of harvest and production. To search for regression models use multiple linear regression methods, while the best models use stepwise methods. Based on existing data, the best model is obtained with negative interception and influenced by productivity and the extent of the yam harvest.