Despite a substantial effect of reference evapotranspiration (ETref) in irrigation practices and hydrological processes, trends and the causes of such trends in ETref are scarcely investigated in Bangladesh. The spatiotemporal trends in ETref, climatic factors influencing the variations in ETref were investigated based on daily climate datasets from 18 sites during 1980-2017. Modified Mann-Kendall, linear regression, Morlet wavelet analysis (MWA) and cross wavelet transform model were employed to appraise temporal variations in ETref and the influential climatic variables. The empirical Bayesian kriging model was used to understand the spatial variations of ETref on the annual and seasonal time scales. The stepwise regression and partial correlation coefficient (PCC) were adopted to identify factors influencing the variations in ETref. The analysis showed a decrease in annual (-1.19 mm/year) and seasonal (-0.40 mm/decade for pre-monsoon, -0.47 mm/decade for post-monsoon, -0.50 mm/decade for winter) ETref except for monsoon in Bangladesh which is similar to “evapotranspiration paradox” observed in some locations. Results of trend analysis also revealed that though a rise in mean temperature (MT), a significant decline in sunshine duration (SD) and wind speed (WS) are the major causes of the decrease in ETref. Spatially, the higher annual ETref was found in the southwestern region while the lower ETref was detected in the northwestern region. The significant periods of 1-3 and 3-5-year cycles were detected in the annual and seasonal ETref. The results exhibited a significant coherence between ETref with climatic variables at various time-frequency bands. Stepwise regression and PCC showed that the effect of climatic variables on ETref differs on the annual and seasonal scales whereas MT, RH, and SD mainly attributed to the variations ETref in Bangladesh. These outcomes are anticipated to be beneficial for irrigation designing a sustainable water practice considering the effects of climate change and anthropogenic contributions.