According to definition of European Environment Agency, drought is a recurrent feature of climate and it cans occur in any range of rainfall, any area and any season (Hughes, 2002, 1572). That is one of the natural dangerous enemies of mankind.Therefore, recognition, review, analysis and forecasting the occurrence of drought is vital for some of human activities. In this study, in addition to analysis of drought (Frequency, Intensity and Strength indices), the repetition of drought spells have been forecasted over Urmia Lake Basin located in the North Western region of Iran. In order to do this, first, the annual precipitations have been spatially normalized in the course of period 1961_2005 to make use BMI Index. After that, the intensity of wet and dry periods was categorized to develop PDRI empirical index in two classes. At the next stage, the time series were applied for drought prediction through SARIMA family models of relevant monthly precipitations forecasting have been implemented by means of SPSS software and according to the most appropriate models for the next 24 months (to 2007). Finally, corresponding the predicted sums, the severity ranks of humid or drought years have been determined. Our study showed that annual precipitation rate is normal approximately with increasing trend over mentioned Basin in 2006-7. On the other hand, the chief of humid and drought years have been interrupted with durability less than a year.