Haors are large, round-shaped floodplain depressions located in the North-Eastern region of Bangladesh. Extreme events such as heavy rainfall routinely affect the haor basin with flash floods. These haors are predicted to experience severe stress because of changes in rainfall and temperature patterns. The biotic community of the wetlands may not have enough time to adjust itself in such varying temperature and rainfall extremes. This paper evaluates various aspects of the future projections of rainfall and temperature extremes, including magnitudes and frequencies thereof. The impacts of extreme events are examined using Hadley Centre’s high-resolution regional climate model known as PRECIS (Providing REgional Climates for Impact Studies). Daily temperature and rainfall simulations of the 17-member ensembles are generated through Hadley Centre Coupled Model (HadCM3). These simulations are used in Rclimdex—a software specially designed for this study. A total of 12 core climate indices are computed, analyzed, and statistically examined (Mann–Whitney U test) over the space of three time slices—(1) short (2020s, i.e., 2011–2040), (2) medium (2050s, i.e., 2041–2070), and (3) long (2080s, i.e., 2071–2098). Here, the 1980s (1971–2000) are considered as the baseline period. The study has found that the highest significant variability in both rainfalls and temperatures was during the pre-monsoon season when flash floods normally occur. Also, rainy days are projected to be less frequent albeit more intense where the deeply flooded haors are located. Though the annual total rainfall does not show any difference in spatial distribution (except for in magnitude), the seasonal patterns of most extreme events show that the probable affected areas have shifted from North-east to further North. In addition, a significant increase in both RX1 (1-day maximum rainfall) and RX5 (5-day maximum rainfall) are projected during the 2080’s pre-monsoon season near Sunamganj. This projection also indicates the possible frequent occurrence of flash floods with high volumes. Probability distribution frequencies (PDF) show a rightward shift in time indicating an increase in the amount of total rainfall in the future. Exceptions are, however, found in case of PDFs for consecutive dry days (CDD) and consecutive wet days (CWD). The decrease in CWD is found to be more pronounced than that of CDD. All these projections made in this study are expected to contribute further in the advancements of the Master Planning of the haor area that was done by the government of Bangladesh in 2012.