Water, a critical component of food, livelihood and economic security, has always received a central place in India’s investment portfolios. The investment in water transfers of the National River Linking Project (NRLP) is one of the biggest proposed in recent times. When and if completed, the NRLP forms a gigantic water grid covering most of South Asia. It envisages transferring 174 billion cubic meters (Bm3) of water across 34 river links and will cost about US$120 billion (2000 prices). The proposed plan has aroused a large interest in recent public discourses. Hydrological feasibility, financial viability and social cost are the issues that dominate these public dalogues. This paper analyzes the cost and benefits of eight river links in the peninsular component, which include the main subcomponent of linking rivers of Mahanadi, Godavari, Pennar and Cauvery. Irrigation is the main beneficiary in this component and, en route, these links or canals account for 85% of the total water transfers to irrigation and domestic and industrial sectors in the command areas. However, our analyses show mixed results of financial viability of individual links. The main reason for this is low net value-added benefits from additional irrigation over and above the existing level of cropping and irrigation patterns. The proposed cropping patterns of these links generate much less net value-added benefits than the existing cropping and irrigation patterns. To make these links financially viable, they need to include high-value cropping patterns that, at least, generate as much benefit per unit area as fruits and vegetables. Although some individual links show less than desirable net benefits, taken together the Mahanadi-Godavari-Pennar-Cauvery subcomponent gives a higher internal rate of return of 14% compared to a discount rate of 12%, and a high benefit-cost ratio of 1.3. However, many unknown factors or unavailable information in this analysis can alter the estimates of financial benefits and costs.