The relationship between soil moisture and outflow is fundamental to land surface water and energy balance monitoring and modeling. However, characterizing it at levels above a point scale is complicated by factors including climate and surface heterogeneity, presence of lateral transports, and mismatches between spatial resolutions of measurements and models. We investigate the distinct roles of heterogeneity and nonlocal interactions in scaling this relationship from points to areas. Locations are modeled as independent columns, using measured soil moisture and precipitation data in a conditional averaging approach to estimate the local moisture outflow relationship. These estimates are aggregated using a distributional approach to account for the effect of heterogeneity. The locations are then treated as one large column; that is, the moisture outflow relationship is estimated directly from spatially aggregated data. We demonstrate that statistically significant differences between the two estimates indicate that the system is not well represented by the independence assumption; that is, local outflow is dependent on local moisture and is also independently influenced by large-scale moisture. We applied these methods to data from a hillslope, a watershed, and the state of Illinois and found that heterogeneity and nonlocal processes significantly affected scaling in all three. The identified nonlocal effect is to decrease (increase) local outflow during large-scale dry (wet) anomalies. A possible pathway is through decreased wind speed during dry anomalies. The combined effect increases the sensitivity of outflow to soil moisture as scale increases. These results could have implications for specifying parameters in large-scale models, especially those calibrated with smaller-scale field data.