Relatively accurate estimation of flood discharge with different return periods in flood control and design projects is essential. The main problem in estimating Dubai is the existence of areas without statistics and with short statistical periods. One of the proposed methods to solve this problem is flood frequency analysis. The first step in analyzing the abundance of flood areas is to determine homogeneous areas, and several methods have been proposed to determine homogeneous areas. In this study, homogeneity method was used to determine homogeneous areas in 3 sub-basins of Gorganroud, Qara-sou and Gorgan Gulf . The tests used in this study include Dalrymple, Hosking and Wallis, Wiltshire, Andrew Curves, Cluster Analysis, and Flood Index Method were examined. The study used data from 28 hydrometric stations that were statistically appropriate. Hydrological, climatic and physiographic characteristics of the basins were used to perform homogeneity tests. In the Dalrymple method, 6 stations were out of the safe range and were excluded from the study as heterogeneous stations. In this method, stations located within the confidence range were divided into two groups. The homogeneity of the stations in the segregated areas was examined using the Husking and Wallis test, in which 3 stations were identified as heterogeneous stations. Wiltshire test, which is divided into two methods of test based on coefficient of variation (S) and test based on distribution (R), showed that the region is heterogeneous based on S statistic and based on R statistic based on the distribution of EV1 homogeneous. In the test, based on the flood index method, 20 stations with a statistical period of 30 years were selected. Out of 20 stations, 4 stations were excluded from the confidence range and 2 stations were removed by Hosking and Wallis methods to confirm homogeneity and finally a lot of analysis frequency. The method of Andrew curves, which is drawn based on trigonometric functions, uses physiographic, climatic and hydrological variables, the region was divided into two groups once based on all variables and again based on the variables obtained from the main components. K-mean and Ward methods were used in the cluster analysis method. In these two methods, which use climatic, physiographic and hydrological variables, once with all the variables and again using the variables obtained from the first components, the region was divided into two groups. In this study, Husking and Wallis test were used to confirm the homogeneity of regions. In most methods, 3 stations of Ploghaznavi-Tilabad, Tangarah and Siahab were identified as heterogeneous stations. Key word: Homogeneous regions, Dalrymple Test, Hosking and Wallis, Cluster Analysis, Wiltshire, Andrew Curves, Flood Index