Time series analysis aims to forcasttime seriesdata in some future period based on the data in the past. The main aim of this research is to forcast the number of the new students of Salafiyah Syafi’iyahSukorejo Boarding School in Situbondo using Auto Regressive Moving Average (ARMA). This research uses annual data from 2005 until 2016. It is discusses the steps of timeseriesanlysis using the Box –Jenkinsmethod. That method comprises of several stages, they are model identification stage, parameter estimation stage, diagnostic checking and forecasting stage. Model identification stage is done by finding the model (p,q) that are considered as the most appropriate by looking at the plot of ACF and PACF of the correlogram. Parameter estimation stage is done by estimating model parameters.Whereas, Diagnostic testing and forecasting stage is done by seeing if residual estimation results is already have the quality of white noise.After the appropriate model has been identified, the next step is to use this model for forecasting. The results of this study shows that the ARMA model (2.0) provide the better forecasting results with squared the smallest value of SSR, AICand SIC.