The aim of the work was to create a new three-dimensional periprosthetic multi-criteria optimisation technique to identify the best six degrees of freedom transform to position a porous-coated anatomic cementless femoral component for three factors, including: first, maximisation of the degree of contact achieved between designated bone ingrowth surfaces and the periprosthetic bone; secondly, minimisation of the bone mass to be removed to accommodate the component and thirdly, the extreme constraint of the component to be positioned so that it does not project beyond the periosteum. Discrete integrals were computed over regions of interest derived from the polyhedral component mesh in transaxial CT scan planes, using a polygon scan-conversion algorithm. A new biomedical imaging volume rendering technique utilising dynamic virtual textures was developed to visualise the design trade-offs. Pareto-optima were identified for four femora that matched an average-sized component. The non-linear, multi-modal fit metric was quadratic near minima, with a narrow trough of equivalent fit values within 3mm of translation and 3 degrees of rotation with respect to the canal axis, and possessed a dependence most pronounced for distal-directed insertion against varus/valgus rotation. The study gives previously unavailable data on the three-dimensional femoral component fit and is the first report that demonstrates that fitting the implant using several design criteria in a multi-criteria optimisation scheme is feasible.