Understanding the effects of spatio-temporal variability on subsurface drainage volumes will help in minimizing the adverse environmental effects on the health of ecological systems. The objectives were to investigate the spatial structure and temporal stability of subsurface drainage trends using six years (1993 to 1998) of field measured data from 36 experimental plots. Two main components of variability (i.e., the large-scale deterministic structure or trend and the small-scale stochastic component) were studied using the median polishing technique and variography. Normalized trend surfaces indicated that trend patterns were stable over the study period. After subtracting the trend from subsurface drainage data, the residuals were used during the subsequent variography. The semivariogram analysis showed a strong spatial structure for most of the years, although the spatial parameters of sill and nugget were found to be different for each year because of climatic effects. The spatial correlation lengths, however, were found to be consistent from year to year at 190 m. Total variance in subsurface drainage data was partitioned between the large-scale deterministic component and the small-scale stochastic component. On average, variations in the trend accounted for about 36% of the total variance, and sill values represented about 64% of the variance. The greater contribution of the stochastic component and stable patterns in the trend surfaces revealed that subsurface drainage flow volumes were controlled by the intrinsic soil and landscape properties. These results indicated that stable trend surfaces can be used as a guide to delineate the agricultural management zones where best management practices can be applied to reduce the negative environmental effects resulting from the discharge of subsurface drainage effluents to surface water bodies such as creeks and streams.