Understanding the basic relationships between rainfall, runoff, soil moisture and ground waterrnlevel are vital for an effective and sustainable water resources planning and managementrnactivities. But so far there are no hydrological studies in Meja watershed that aims to understandrnthe watershed characteristics and runoff generation processes. This study was conducted tornunderstand runoff generation processes and model rainfall runoff relationship in Mejarnwatershed having a drainage area of 96.6 km2.The watershed is one of the three research sites ofrnInternational Water Management Institute (IWMI) developed in early 2010 in the upper BluernNile Basin Ethiopia. In the study, primary data of soil moisture, shallow ground water level,rnrainfall and runoff were collected from the hydrological monitoring network in the watershed.rnTwo nested sub- watersheds namely Galessa and Kolu were defined in the watershed for detailrnanalysis of hydrologic variables. Galessa has drainage area of 1.6 km2 and Kolu has a drainagernarea of 2.5 km2. Hydrological models like HBV and RRL SMAR were configured to understandrnthe relationship between rainfall and runoff in the watershed. Relationships between rainfall,rnsoil moisture, shallow ground water level and runoff were developed to understand runoffrngeneration processes in the watershed. Analysis of rainfall data indicated weak daily correlationrn(r20.8). According to one yearrnand three months data, there is no strong daily rainfall and runoff relationship (r20.65) in monthly basis of Kolu nested sub watershed but there is moderate relationship ofrnrainfall and ground water level. There is spatial variability of soil moisture content in Mejarnwatershed, this variation occurs due to heterogeneity of the soil, which means the places arerndifferent in soil texture and also the variation is due to vegetation cover and change of slope.rnThere is strong linear relationship of rainfall and monthly averaged volumetric soil moisture inrnmost soil moisture layers of Meja and its nested sub-watersheds. The general relationshiprnbetween runoff and monthly averaged soil moisture at different layers in Meja watershed andrnKolu is strong. Analysis of rainfall runoff models indicated that relationships of rainfall with observed and simulated runoff was similar.HBV model performs better than RRL SMAR model inrnMeja and Kolu. RRL SMAR model couldn’t capture low flow in Meja and Kolu. This inaccuraternresult of SMAR model in Kolu sub-watershed may be due to inability of the model to simulaternrunoff in very small catchments like Kolu.