Bangladesh is one of the most disaster-prone countries in the world, with great negative consequences being associated with various natural and human-induced hazards. Therefore, the evaluation of long-term variation in recorded meteorological and hydrological time-series data is necessary for infrastructure development in the future. In this study, we collected 32-year monthly precipitation data at 36 locations in Bangladesh, 116-year monthly precipitation data at seven areas in India and 30-year data of the daily water level at four locations in the Ganges Brahmaputra Meghna and Haor in Bangladesh. We conducted a trend analysis using these data as well as a correlation and causality analysis in order to find the relation to global-scale phenomena such as Southern Oscillation Index (SOI) and Dipole Mode Index (DMI). We used the Mann-Kendall test and Seasonal Mann-Kendall test for long-term trend test of precipitation and river water level; the wavelet and cross-wavelet coherence analysis for correlation; and VAR-LiNGAM for causality between the precipitation and SOI, the precipitation and DMI. The Mann-Kendall test and Seasonal Mann-Kendall test showed a significant decrease in precipitation in the central part of Bangladesh. Meanwhile, no significant decrease is shown in the northern part and no trend of precipitation is shown in other parts. The slope of the 32-year trend is found to be bigger compared to the 116-year data of India. The daily water level of three major rivers show a significant decrease in the recent 30 years. Especially, the slope of the trend in Brahmaputra is bigger than those of the other parts of Bangladesh. For the relation to global-scale phenomena, the cross-wavelet coherence analysis shows that the correlation of DMI to monthly precipitation is stronger than that of SOI. The VAR-LiNGAM analysis shows the causality from DMI to precipitation in the Bengal area with five to six months lag.