Assessing risk associated with computational grid is an essential need for both the resource providers and the users who runs applications in grid environments. In this chapter, we modeled the prediction process of risk assessment (RA) in grid computing utilizing meta-learning approaches in order to improve the performance of the individual predictive models. In this chapter, four algorithms were selected as base classifiers, namely isotonic regression, instance base knowledge (IBK), randomizable filtered classified tree, and extra tree. Two meta-schemes, known as voting and multi schemes, were adopted to perform an ensemble risk prediction model in order to have better performance. The combination of prediction models was compared based on root mean-squared error (RMSE) to find out the best suitable algorithm. The performance of the prediction models is measured using percentage split. Experiments and assessments of these methods are performed using nine datasets for grid computing risk factors. Empirical results illustrate that the prediction performance is enhanced by predictive model using ensemble methods.