Bangladesh is one of the most climatic hazard-prone countries in the world. Rainfall in Bangladesh is highly variable which depends on each season and location. This study aims to investigate spatiotemporal variability of rainfall linked to groundwater level under changing climate in the northwestern region, Bangladesh during the period of 1976-2016 using rainfall and groundwater level data. A number of statistical approaches such as Mann-Kendall (MK) test, linear regression model, Sen’s slope estimator and Spearman’s rho (SK) test, Coefficient of Variation (CV) were employed to reveal trends and variability of rainfall and groundwater level. Additionally, wavelet transform was applied to assess temporal variability of rainfall time series in the northwestern region, Bangladesh. The results show that the annual rainfall variability was the highest in the Ishwardi station (25.39%) and the lowest was found in Rajshahi (20.70%). The MK test results reveal that the trends of annual and seasonal rainfall decreased gradually, where the highest significant decreasing trend of annual rainfall was found in Rajshahi (-2.976 mm/year) and the lowest decreasing trend was found in Syedpur (-1.278 mm/year). Furthermore, monsoon and winter seasons represent the decreasing trends of rainfall for all the stations. The results of wavelet analysis demonstrate that the monthly rainfall variance of average rainfall in the northwestern region was low from the period of 2006 to 2014, indicating the dry years. The annual and seasonal rainfall trends in the northwestern region were decreased, particularly in the recent decade. The relationship between rainfall and groundwater level indicates that rainfall was decreased and the ground water level was also declining at the same time in the study region. Interestingly, it is observed that a significant high decreasing trend of annual rainfall is found in Rangpur but the annual groundwater level depth is less increased in Rangpur at the same time.