The soybean aphid, Aphis glycines Matsumura (Hemiptera: Aphididae), is capable of reducing soybean, Glycine max (L.) Merr., yield up to 40% during severe outbreaks. Frequent sampling, which can be costly and time-consuming, is essential to making informed management decisions. However, one way to decrease sampling effort is to use a reduced sample unit when possible. The objectives of this study were to describe the vertical distribution of soybean aphid within soybean over time and to define node-based sample units of varying sizes by testing the ability of selected units to accurately estimate whole-plant aphid density. Within-plant distribution of soybean aphid changed significantly with time. However, the average nodal position where soybean aphids were found on soybean remained within the top half of the plant at all three locations studied across all sample dates. Consequently, selecting the node with the highest aphid density multiplied by the total number of infested nodes (N(MAX)') was the best predictor of aphids on remaining soybean components in both the original (r2 = 0.855) and validation (r2 = 0.824) data sets. For sample units that included more than a single node to estimate densities, a weighted formula, which incorporated changes observed in the within-plant aphid distribution, improved model performance (higher r2 values) and reduced variability around parameter estimates compared with a node-averaged formula. Our results suggest that smaller sample units provide reliable estimations of whole-plant aphid density throughout the growing season for differently maturing soybean, which is essential to their use in pest management decisions and development of future sampling plans.