Variations in weather and climate have a significant impact on rain-fed banana yields in East Africa. This study examined empirical linkages between banana yields and variations in rainfall and temperature over Uganda for the historical period (1971-2009) using time series moments, correlation and regression analysis. The Food and Agriculture Organization (FAO) Crop Water Assessment Tool (CROPWAT) was used to estimate banana crop water requirements, soil moisture deficits and their effects on banana yield levels under rain-fed conditions for different regions. The study observed high comparability in moment indices with some significant differences reflected in the values of the banana yields and rainfall and temperature moment indices. The cumulative effect of rainfall and temperature variations on banana yields was discernible from strong correlation coefficients of up to 78%. The CROPWAT simulations indicated up to 46% reductions in optimal banana yields due to soil moisture deficits within banana plantations. In conclusion, the study observed stronger linkages between banana yields and temperature variations than rainfall. In addition, temperature manifests both direct and indirect effects on banana growth while rainfall exhibits comparatively high intra-seasonal and intra-annual variability with lag effects on banana yields. The study provides a strong scientific basis for the development of coping, adaptation and mitigation strategies in the banana farming subsector in the region due to the anticipated shifts in rainfall and temperature extremes and changes across Uganda and neighbouring regions.