Crop yield variations in the Orinoquía region, Colombia, are primarily associated with extreme precipitation events. Therefore, studying crop water supplies under naturally variable climate conditions is fundamental in an actual climatic change context. Rainfall data collected in the Quenane basin were analyzed to understand the soil water dynamics in the Orinoco catchment. The basin covers 179 km2 and consists of the piedmont landscape (Eastern Mountain Range) of the Villavicencio Municipality, Department of Meta. This study analyzes the rainfall variability using Pearson correlation analysis, the Mann–Kendall trend analysis, and soil water balance to determine the implications of these factors in crop performance at the basin scale. The results indicated that the spatial distribution of rainfall in the basin responds to a longitudinal average variation of precipitation and that this response is more accentuated (i.e., greater rainfall) toward the west of the basin. Despite the basin being located in the tropical zone, no evidence was found regarding the effect of the El Niño Southern Oscillation on rainfall patterns. Yet, the temporal analysis revealed some years with extreme rainfall values and high-uncertainty levels during transitions between wet and dry periods. During these transition periods, a greater potential for effects on farm yields exists due to the variable cumulative rainfall observed during recent years. The time series trend analysis revealed changes in rainfall patterns at different scales (weekly and yearly) and distribution based on the decrease of rainy days per week and year. This trend is much more accentuated during the second half of the year, generating uncertainty and reducing farm yields throughout the basin.