Since Solar Flares are associated with the sunspots. Solar flares have different length, duration and peakness in different intervals as sunspots have. The Solar Flares index data starting from 1966 to 2008 (Monthly Flare Index Cycles" 20, 21, 22, 23") in both North and South hemisphere separately along with total Solar Flares data. This paper compared the North, South and total hemispheric Solar Flares data in perspective of probability distributions such as Gamma, Log-Gamma and Chi-square which tested by Kolmogrove-Smirnov D-test and then their persistency by using fractal dimension. Further we estimate the Hurst exponent with the help of fractal dimension. Fractal dimension shows complex of the data while Hurst exponent represents smoothness of the data. All the tested probability distributions in this paper show persistency and positively correlated. The use of probability distributions allows to represents the uncertainties in the solar flares data. This study will be helpful to verify the long term smoothness of data trends and conclude the results will be present so forecast values will be more accurate.. Keywords: Solar Flares, Gamma Distribution, Log-Gamma Distribution, Chi-Squared Distribution, Fractal Dimension, Hurst Exponent