The sediment hyper concentration in the lower Yellow River makes it difficult to simulate sediment concentration accurately when using traditional numerical model. This paper uses a variational based data assimilation method to improve the result of simulated sediment concentration in the lower Yellow River, China. A cost function is introduced to describe difference between simulated sediment concentration and observations. One dimensional suspended sediment transport equation, taken as a constraint, is added into the cost function. Adjoint equation of the data assimilation system is used to solve the minimum problem. The usefulness and the capability of the variational data assimilation are demonstrated in the lower Yellow River during the water-sediment regulation in 2011. In the assimilation process, errors between the simulations and the observations are analyzed. It is believed that the flexible approach of the variational data assimilation can improve the accuracy of simulated sediment concentration. Finally, limitations of variational data assimilation and future work are discussed.