In a rapidly changing water resources system, dynamic models based on the notion of systems thinking can serve as useful analytical tools for scientists and policy-makers to study changes in key system variables over time. In this paper, an integrated system dynamics simulation model was developed using a system dynamics modelling approach to examine the feedback processes and interaction between the population, the water resource, and the agricultural production sub-sectors of the Volta River Basin in West Africa. The objective of the model is to provide a learning tool for policy-makers to improve their understanding of the long-term dynamic behaviour of the basin, and as a decision support tool for exploring plausible policy scenarios necessary for sustainable water resource management and agricultural development. Structural and behavioural pattern tests, and statistical test were used to evaluate and validate the performance of the model. The results showed that the simulated outputs agreed well with the observed reality of the system. A sensitivity analysis also indicated that the model is reliable and robust to uncertainties in the major parameters. Results of the business as usual scenario showed that total population, agricultural, domestic, and industrial water demands will continue to increase over the simulated period. Besides business as usual, three additional policy scenarios were simulated to assess their impact on water demands, crop yield, and net-farm income. These were the development of the water infrastructure (scenario 1), cropland expansion (scenario 2) and dry conditions (scenario 3). The results showed that scenario 1 would provide the maximum benefit to people living in the basin. Overall, the model results could help inform planning and investment decisions within the basin to enhance food security, livelihoods development, socio-economic growth, and sustainable management of natural resources.