A review has been made to understand the hydrogeochemical behaviour of groundwater through statistical analysis of long term water quality data (year 2005–2013). Water Quality Index (WQI), descriptive statistics, Hurst exponent, fractal dimension and predictability index were estimated for each water parameter. WQI results showed that majority of samples fall in moderate category during 2005–2013, but monitoring site four falls under severe category (water unfit for domestic use). Brownian time series behaviour (a true random walk nature) exists between calcium (Ca2+) and electric conductivity (EC); magnesium (Mg2+) with EC; sodium (Na+) with EC; sulphate (SO2−4) with EC; total dissolved solids (TDS) with chloride (Cl−) during pre- (2005–2013) and post- (2006–2013) monsoon season. These parameters have a closer value of Hurst exponent (H) with Brownian time series behaviour condition (H=0.5). The result of times series analysis of water quality data shows a persistent behaviour (a positive autocorrelation) that has played a role between Cl− and Mg2+, Cl− and Ca2+, TDS and Na+, TDS and SO2−4, TDS and Ca2+ in pre- and post-monsoon time series because of the higher value of H (>1). Whereas an anti-persistent behaviour (or negative autocorrelation) was found between Cl− and EC, TDS and EC during pre- and post-monsoon due to low value of H. The work outline shows that the groundwater of few areas needs treatment before direct consumption, and it also needs to be protected from contamination.