Background During implantation of an uncemented femoral knee implant, press-fit interference fit provides the primary stability. It is assumed that during implantation a combination of elastic and plastic deformation and abrasion of the bone will occur, but little is known about what happens at the bone-implant interface and how much press-fit interference fit is eventually achieved. Methods Five cadaveric femora were prepared and implantation was performed by an experienced surgeon. Micro-CT- and conventional CT-scans were obtained pre- and post-implantation for geometrical measurements and to measure bone mineral density. Additionally, the position of the implant with respect to the bone was determined by optical scanning of the reconstructions. By measuring the differences in surface geometry, assessments were made of the cutting error, the actual interference fit, the amount of bone damage, and the effective interference fit. Findings Our analysis showed an average cutting error of 0.67 mm (SD 0.17 mm), which pointed mostly towards bone under-resections. We found an average actual AP interference fit of 1.48 mm (SD 0.27 mm), which was close to the nominal value of 1.5 mm. Interpretation We observed combinations of bone damage and elastic deformation in all bone specimens, which showed a trend to be related with bone density. Higher bone density tended to lead to lower bone damage and higher elastic deformation. The results of the current study indicate different factors that interact while implanting an uncemented femoral knee component. This knowledge can be used to fine-tune design criteria of femoral components to achieve adequate primary stability for all patients.