Although mechanized harvesting of sugarcane can benefit the soil–environment system, successive harvests can compromise the physical soil quality. Information on the growth and distribution of the sugarcane root system and its relationship with physical soil properties are scarce in the literature. The objective of this study was to analyze the relationships between the spatial distribution of physical soil properties and sugarcane root density in a mechanically harvested area. The study was carried out in Jaboticabal, SP, Brazil in a Typic Kandiudox and with the SP81 3250 cultivar. Sugarcane plots were established in April 2005, and samples were taken from four trenches in July 2008 after three successive harvests. In each trench, 15 undisturbed soil samples were collected to evaluate bulk density ( Ds), penetration resistance ( Pr), total porosity ( Tp) and moisture at field capacity ( θ fc ), and 15 monoliths were collected to evaluate the root density ( Rd), root length ( Rl), root area ( Ra) and root moisture ( Rm) at different depths and plant locations. Soil moisture was balanced to the field capacity before Pr determination. The hypothesis that the sugarcane root system varied vertically and horizontally relative to the central axis of the plant and in relation to physical soil parameters was validated. Soil physical parameters varied significantly in the vertical direction, but sugarcane root system varied both vertically and horizontally. Ds values were 1.57, 1.55 and 1.39 g cm −3 for layers between 0 and 0.2 m, 0.2 and 0.4 m and 0.4 and 0.6 m, respectively. For the same layers, the Pr was 1.42, 1.20 and 0.87 MPa. The largest Pr value (1.93 MPa) was obtained near the surface in the space between rows. As Ds increased, Tp decreased and Pr increased. A consistent relationship was observed between Ds and Pr in all layers, and restrictions on root growth ( Pr > 2.0 MPa) may occur when Ds ≥ 1.78 g cm −3. The multivariate boundary line approach, in combination with nonlinear multiple regression analysis, successfully quantified the effects of Pr on Rd, Rl and Ra. Sugarcane root growth was not affected below Pr values of 0.75 MPa, but decreased significantly between 0.75 and 2.0 MPa. Root growth was severely restricted when Pr > 2.0 MPa.