As-cast, porous surfaced CoCr implants were tested for bone interfacial shear strength in a canine transcortical model. Three-dimensional printing (3DP) was used to create complex molds with a dimensional resolution of 175 microm. 3DP is a solid freeform fabrication technique that can generate ceramic pieces by printing binder onto a bed of ceramic powder. A printhead is rastered across the powder, building a monolithic mold, layer by layer. Using these 3DP molds, surfaces can be textured "as-cast," eliminating the need for additional processing as with commercially available sintered beads or wire mesh surfaces. Three experimental textures were fabricated, each consisting of a surface layer and deep layer with distinct individual porosities. The surface layer ranged from a porosity of 38% (Surface Y) to 67% (Surface Z), whereas the deep layer ranged from 39% (Surface Z) to 63% (Surface Y). An intermediate texture was fabricated that consisted of 43% porosity in both surface and deep layers (Surface X). Control surfaces were commercial sintered beaded coatings with a nominal porosity of 37%. A well-documented canine transcortical implant model was utilized to evaluate these experimental surfaces. In this model, five cylindrical implants were placed in transverse bicortical defects in each femur of purpose bred coonhounds. A Latin Square technique was used to randomize the experimental implants left to right and proximal to distal within a given animal and among animals. Each experimental site was paired with a porous coated control site located at the same level in the contralateral limb. Thus, for each of the three time periods (6, 12, and 26 weeks) five dogs were utilized, yielding a total of 24 experimental sites and 24 matched pair control sites. At each time period, mechanical push-out tests were used to evaluate interfacial shear strength. Other specimens were subjected to histomorphometric analysis. Macrotexture Z, with the highest surface porosity, failed at a significantly higher shear stress (p = 0.05) than the porous coated controls at 26 weeks. It is postulated that an increased volume of ingrown bone, resulting from a combination of high surface porosity and a high percentage of ingrowth, was responsible for the observed improvement in strength. Macrotextures X and Y also had significantly greater bone ingrowth than the controls (p = 0.05 at 26 weeks), and displayed, on average, greater interfacial shear strengths than controls, although they were not statistically significant.