Abstract
This paper presents application of the method for assessment of predictive uncertainty of a rainfall‐runoff model using data‐driven modelling techniques. Historical model errors which are mismatch between observed and simulated values of runoff by rainfall‐runoff models are assumed to be indicators of total model uncertainty. Uncertainty is represented by the model prediction interval. Different data‐driven techniques used to determine the interval are compared. The method is tested to estimate uncertainty of runoff simulations by conceptual rainfall‐runoff model with application to Brue catchment in the United Kingdom; it is also compared with other uncertainty estimation methods.