The intensity duration frequency relationship of rainfall amount is one of the most commonly used tools in water resources management for planning, designing and operation of water resources projects. This study aimed to develop an effective and relatively simple regional IDF curves development for the upper Blue Nile Basin in Amhara regional state. Four sub hourly and 58 daily recording stations were collected from different sources for this study and identification of homogeneous region was performed for development of regional IDF curves for each region. Tests for gross outliers and inconsistencies were done to check the appropriateness of the data. The mean annual maximum rainfall is normalized for all rainfall durations that used for frequency analysis. To estimate a chosen distribution parameters and obtain quantile estimates, the regional L-moment algorithm was used. This procedure involves fitting the chosen distribution using the method of L-moment; its parameters were estimated by equating the population L –moments of the distribution to the sample L-moments derived from the observed data. Next, sample L-moments ratios from each site in a homogeneous region were weighted, according to record length, and combined to give regional average L–moment ratios. Regional parameters and quantile were estimated based on the selected distribution and quantile estimation at each site i is obtained by multiplying the regional quantile by the mean of maximum annual daily rainfall at site i. The result shows that the basin is grouped into two homogenous regions and it was found that Pearson type-III distribution best fitting for the rainfall duration of 0.5hr, 1hr, 3hr and 24hr for region one and 0.5hr and 1hr for region two. Extreme Value Type-I distribution best fitting for 2hr and 6hr rainfall duration for region one, 12hr and 24hr for region two. Generalized Pareto distributions are the best fit for 12hr duration for region one and Generalized Logistic distributions best fit for 2hr,3hr and 6hr rainfall duration for region two. The existing IDF in the study area, overestimate areas which have low rainfall intensity with an average of 37% times than expected rainfall intensities and under estimate for areas having high rainfall intensities with an average of 36%times than the expected rainfall intensities.