Sensitivity analysis is an important component of water resource and environmental modelling. Analysis of Variance (ANOVA), as global sensitivity analysis technique, has been widely used for achieving this. To diminish the effect of the biased variance estimator of ANOVA, three developed subsampling (single-, multiple- and full-subsampling) ANOVA approaches are established. Two case studies including one simplified regression model and one hydrological model are used to illustrate the performance of these approaches. The traditional Sobol’s method is used as benchmark method. Results find that: (1) The subsampling effectively diminishes the bias introduced by the biased variance estimator. (2) The difference of sampling densities among parameters has great influence on quantification of parametric sensitivities in hydrologic modeling. (3) Compared with Sobol’s method, the subsampling ANOVA methods can significantly reduce the calculation requirements while achieve similar calculation accuracy. This study serves as a first basis for the application of subsampling ANOVA in water resource and environmental models sensitivity analysis.