The aim of this paper is to deal with one-way analysis of variance (ANOVA) in fuzzy environment. The traditional method of ANOVA has been generalized based on fuzzy random variables instead of real-valued random variables. All random errors and parameters are considered as triangular or exponential fuzzy numbers. The least squares method is employed to estimate the fuzzy parameters and a method is discussed to check the adequacy of the proposed model. Finally, two examples are provided to clarify the presented discussion in this article.