Floods are generally associated with the calamitous impacts in the country like India. For the assessments of flood peak magnitude, the most common method adopted is the flood frequency analysis. In this case study the four common methods namely Gumbel Max, Generalized Extreme Value (GEV), Log Pearson type 3 (LP3) and Log10 Normal distribution are adopted to analyze the flood frequency and flood recurrence for the Mahanadi River. River discharge values from 1979 to 2018 (40 years) for different gauging stations namely Ghatora, Rajim, Kesinga, Kotni, Kantamal and Jondhra located in the basin are analyzed. Two goodness-of-fit tests namely Kolmogorov–Smirnov and Chi-squared are employed for appropriate probability distribution modelling of flood at 5% level of significance. The outcomes reveal that the Log Pearson type 3 (LP3) is the best fit for Rajim, Jondhra, Kantamal, Kesinga and Kotni monitoring stations and Log10 Normal is found to be the best fit for the Ghatora monitoring station. Importance of the current study lies in predicting the design flow values based on the recurrence interval after analyzing with the applicable distribution for the monitoring station concerned.