This study aims to predict the yield of chili plants in Indonesia using multiple linear regression method. In this study, the variables analyzed include irrigation volume, temperature, soil moisture, soil pH, and plant growth parameters such as stems, branches, and leaves. Data were collected from 100 chili plant samples planted in Jatikusuma Village, Deli Serdang Regency, for 63 days. The method used for the analysis is multiple linear regression, which is applied to produce a prediction model of harvest yield. Multi linear regression method is used to perform forecasting with the development of the dependent variable (Y), namely the amount of production with independent variables consisting of x1 = plant growth rate, x2 = moisture, x3 = temperature, x4 = volume, x5 = soil pH, x6 = stem, x7 = branch, x8 = leaf. The results of the prediction analysis in this study obtained the intercept coefficient value is 153.94 from the total data of 100 samples, resulting in the level of fit of the multi linear regression model with an R2 score of 1.00 which shows the level of accuracy in a prediction of these results is very good.