Among deciduous fruit crops, maximum sustained yields under ideal conditions reportedly range from 22 MT/ha in sweet cherry to 112 MT/ha in apple: about a 5-fold difference. Root length densities under fruit trees, however, range from about 0.2 km/m2 in apple to about 12 km/m2 in kiwifruit: a 60-fold difference. What causes these differences among the root systems of different fruit crops? Patterns of root growth and distribution are reviewed as well as recent work on root pigmentation and survivorship. These data illustrate the importance of recognizing fine roots as a heterogeneous population with different longevities and physiological functions. Both root costs and benefits need to be understood to evaluate root efficiency. The concept of root efficiency has been largely ignored in horticulture, despite the considerable cost of building and maintaining the root system. Plants expend carbon (photosynthate) on constructing new roots and maintaining existing roots. Root efficiency can be defined as the ratio of water or nutrient benefit to carbon cost over the lifetime of the root. Apple roots are readily shed when they become inefficient, as occurs when soil temperatures are elevated, soil becomes dry, or when roots are located in infertile patches of soil. For example, in a field experiment with 'Red Chief Delicious' on M.26 rootstock, only 23% of apple roots born in June were still alive in September when the soil was unirrigated and heated, whereas 45% were still alive in the irrigated and unheated treatment. In a split-pot study in the greenhouse, apple seedlings whose roots received 8 mM nitrate lived about 50% longer than its portion of roots that only received 1.6 mM nitrate. High root efficiency may partly explain why apple trees are able to produce relatively high yields despite their sparse root system.