The influence of stem size and extent of porous coating on femoral bone resorption was examined in 411 cases of primary cementless hip arthroplasty. Moore design, cobalt alloy femoral implants with powder-made sintered porous coating on either one-third, two-thirds, or the full implant length were compared radiographically two years after surgery. A semiquantitative method was adopted for assessing resorption that involved dividing the anteroposterior (AP) and lateral roentgenograms into a total of 16 discrete sites. The 16 sites were qualitatively examined for evidence of resorption by either thinning or darkening of bone relative to the time immediately following surgery. Based on the number of sites that demonstrated resorption, the bone loss was classified as either minor and not likely to cause problems (0 to 4 sites) or pronounced and of potentially harmful clinical consequence (5 or more sites). Pronounced resorption occurred in 18% of the 411 cases. The use of larger stems resulted in increased occurrence of marked bone resorption: stems greater than or equal to 13.5 mm in diameter showed five times the incidence of pronounced resorption compared with stems less than or equal to 12.0 mm in diameter. Stems with two-thirds and full porous coating resulted in a twofold to fourfold increase in the incidence of pronounced bone resorption. The theoretic degree of stress shielding of the femoral shaft in bending was calculated for cases with complete canal filling and a radiographic appearance of bone ingrowth. There was a strong correlation between this theoretic factor and the observed bone resorption.