Computational grid is an aggregation of geographically distributed network of computing nodes specially designed for compute intensive applications. The diversity of computational grid helps in resource utilization in order to support execution of all types of jobs; fine grain as well as coarse grain. It is observed that, over the period of time in the course of job execution, grid becomes highly imbalance resulting in performance degradation. It warrants balancing the load amongst the grid nodes. In absence of centralized information in a system such as grid, load balancing becomes a complex problem. Genetic Algorithm, a search procedure based on evolutionary computation, is able to solve a class of complex optimization problems. A model based on genetic algorithm is proposed, in this work, to achieve better load balancing in computational grid. To study the performance of the proposed model, experiments have been conducted by simulating the model. Experimental results reveal the effectiveness of the proposed model.