abstract The study aimed to predict the production of the tomato crop in its three plantings using time series for the production of the tomato crop during the period of (1999- 2014). The test of ADF showed unit root and Normality test showed that the time series data of the study does not significantly differ from the normal distribution of the random variable .The use of modern analysis procedures based on Box-Jenkins methodology showed efficient prediction of tomato crop production. The model of ARIMA (0,0,1) was the best model among the proposed models in this study to predict the winter tomato crop where production is forecasted at 4165, 4238, 4312, 4385 and 4458 thousand tons during the period of (2016- 2020), respectively. The proposed model to predict summer tomato crop is the model ARIMA (2,0,1), where production is forecasted at 4801, 5058, 5209, 5258 and 5246 thousand tons during the period of (2016- 2020), respectively. The model of ARIMA (0,0,1) was proposed to predict the indigo tomato crop where production is estimated at 769, 747, 726, 705 and 684 thousand tons during the period of (2016- 2020), respectively.The study recommends using models that have been reached to predict the production of the tomato crop and the adoption of forecasts given by the proposed models to put the future plans to develop the tomato manufacturing and to export it processed and fresh. Using methodology of Box-Jenkins in the conclusion and the development of standard models to predict other crops, according to the actual development of the time series subjected to the study.