Root systems are important for global models of below-ground carbon and nutrient cycling. Notoriously difficult sampling methods and the fractal distribution of root diameters in the soil make data being used in these models especially susceptible to error resulting from under-sampling. We applied the concept of species accumulation curves to root data to quantify the extent of under-sampling inherent to minirhizotron and soil coring sampling for both root uptake and carbon content studies. Based on differences in sample size alone, minirhizotron sampling missed approximately one third of the root diameters observed by soil core sampling. Sample volumes needed to encounter 90% of root diameters averaged 2481 cm3 for uptake studies and 5878 cm3 for root carbon content studies. These results show that small sample volumes encounter a non-representative sample of the overall root pool, and provide future guidelines for determining optimal sample volumes in root studies.