The arid and semi-arid regions with water scarcity are vulnerable to several stressors such as urbanization, high water demand created by agricultural and industrial activities, point and non-point pollution sources, and climate change. Hence, proactive policies and sustainable water management strategies that are based on decision support systems are crucial in arid and semi-arid regions. Because of large expenses and implementation difficulties associated with the diffuse pollution abatement plans, many authorities are hesitant to initiate, especially those that may present a financial burden on population. Lake Mogan, a shallow lake, is located in a semi-arid region dominated by dry agricultural activities and has been in eutrophic state for the past 20 years. There has been several management alternatives suggested to improve the water quality in Lake Mogan and one of the alternative is the application of BMPs that include fertilizer management, conservation/no tillage, contouring, and terracing to reduce the amount of diffuse source pollutants. In this study, Soil and Water Assessment Tool (SWAT) Model is applied to evaluate the effectiveness of agricultural best management practices (BMPs) in the Lake Mogan watershed located in a semi-arid region. The most effective BMP scenario was found as the one in which three individual BMP scenarios (30% fertilizer reduction, no tillage, and terracing) were combined. With this scenario average annual load reductions of 9.3%, 8.6%, 8.0%, and 11.1% were achieved in sediment, nitrate, total nitrogen, and total phosphorus, respectively. Even with the most effective BMP strategy, high levels of nutrient reduction will not be achieved since non-irrigated agriculture and intermittent low-flow streams accounts majority of the study area. The outcomes suggest integrated solutions should be developed to improve water quality in Lake Mogan. It is aimed that this study will aid decision makers to implement effective best management practices in watersheds showing similar characteristics (i.e. topographical, hydrologic processes, LULC (Land use land cover) characteristics, agricultural activities, meteorological etc.) with the study area. (C) 2017 Elsevier B.V. All rights reserved.