Choice of a design rainfall is a basic step in many commonly used methods for estimation of design floods. Determination of rainfall temporal pattern is of particular importance to estimate flood specially for urban and rural floods, design of runoff excrement systems, culverts and drains, and also soil erosion studies and estimation of storms flooding potential. In this research, 9 rain gauges’ graphs were used to determine rainfall temporal pattern in Sistan and Baluchestan province scattered throughout the province (From Meteorology organization and Energy ministry for scince 1968-2006 years). All intensive storms for each station were studied with 1-hour, 2-hour, 3-hour, 6-hour, 9-hour, 12-hour, 18-hour and 24-hour durations (1136 storms in addition). The method used was based on storms separation into four quarters and estimation of rainfall percentage for each quartile and quartile ranking, then ranks averaging in each quartile, specifying index rank and at last, obtaining final pattern for each duration.To statistically evaluate obtained patterns with durations of interest, Chi square test was used after establishing a contingency table. Results show that in all stations, maximum 1-hour and 2-hour precipitation occur in first and second quartiles and first quartile, respectively. In addition, in Sistan and Baluchestan province with hot and dry climate, almost 80 percent of short-term (1, 2, 3, 6-hour) precipitation takes place in first and second quartiles while long-term (9, 12, 18, 24-hour) precipitation occurs in third quartile.