Competition affects increases in production and economic growth. Good production planning has direct effects on production system modifications. The company in the case study was an Installation Elevator Company currently encountering problems in the area of aggregate production planning and sales prediction, causing the company to be unable to deliver products on schedule with effects on customer satisfaction. This study had the objective to study and compare prediction methods of companies in the Installation Elevator Company and to study aggregate production planning in order to analyze and determine the lowest aggregate production cost by using data from 12 months in retrospect to determine the most appropriate prediction method. This study used 6 prediction methods consisting of the Moving Average Method, the Simple Exponential Method, the Double Exponential Method, the Linear Trend Method, the Quadratic Trend Method, and the Winter Model. Accuracy of predictions was measured using the lowest mean absolute percent error (MAPE). According to the findings, the prediction method with the most appropriate model was the Quadratic Trend Method because the method had the lowest mean absolute percent error of 17.85. Scores from predictions were then used to make aggregate production plans processed through Lingo Program Version 13 under changing circumstances of demand in 5 cases. The findings revealed the lowest production cost for normal demand was 3,233,000 baht and 2,607,000 baht when demand declined by 20 percent.