The Generalized Extreme Value (GEV) distribution was identified as a good model for flood frequency analysis in hydrology. In this study, annual daily maximum flood heights data from 1970 to 2017 were modeled for five hydrometric sites in the Mahanadi River Basin, first time using the q-Generalized Extreme Value distribution (q-GEV) over GEV distribution model. The target of the study was met by estimating the parameters of the distributions using method of maximum likelihood estimation and performing Kolmogorov-Smirnov, Anderson-Darling and Cramer-von Mises goodness of fit tests; information based criteria AIC and BIC for q-GEV distribution and made comparison with GEV distribution. The simulation study was also conducted for checking the suitability of the model. The results revealed that q-GEV distribution performed better than GEV distribution in modeling the extreme events. This model is more helpful to water practitioners for predicting the extreme events and in taking necessary preparations to mitigate the bad effects of flooding on livings, crops and assets in the associated region.