Urban land development frequently destroys soil structure and removes organic matter, limiting tree growth. Soil rehabilitation has potential to improve soil quality but the long-term effectiveness and consequences for tree growth are poorly documented. We evaluated growth, canopy development, and physiological response of five tree species over six years to soil rehabilitation in an experimental site pre-treated to replicate typical land development. A corollary experiment evaluated growth and establishment of three additional species one year after rehabilitation in highly urbanized sites in Arlington County, Virginia. Plot study soil treatments were: typical practice (TP) (10cm topsoil replaced); enhanced topsoil (ET) (topsoil+rototilling); profile rebuilding (SPR) (compost amendment via subsoiling to 60-cm depth+topsoil+rototilling); and undisturbed (UN) (agricultural land with no pre-treatment). In Arlington, SPR was compared with conventional site preparation (topsoil replacement). Overall, trees grew more rapidly in SPR soils and soil depths immediately below the surface (∼15–30cm) were most affected by SPR, which reduced soil bulk density by between 0.19 and 0.57Mgm−3 compared to nonrehabilitated soils. After six years, both trunk cross-sectional area and canopy area of plot-study trees in SPR soils matched or surpassed those in undisturbed soil for all species except Quercus bicolor while canopy area increased by as little as 2% (Q. bicolor) to as much as 84% (U. ‘Morton’). In Arlington, SPR resulted in 77% trunk cross-sectional area growth after one year. Plant and soil water relations may also be altered by rehabilitation, possibly contributing to its potential as a tool for stormwater mitigation. Rehabilitation accelerates establishment and growth of urban trees planted in compacted urban soils indicating that the below-ground environment should be a key component in policy and decision making.