As implant survival extends into the second and third decades focal osteolysis around cemented femoral components in total hip replacement is emerging as an important failure mechanism. Whilst the problem of focal osteolysis is well recognised, there are many aspects of its development which are poorly understood. The broad aim of this thesis is to try to provide some insights into how, why and where focal osteolysis develops around the cemented femoral component. There are broadly two sections to this thesis, chapters 2-5 present clinical and geometrical studies and chapters 6-10 a series of experimental studies. The aim of the first section was to establish what is observed in clinical practice, the aim of the second to try to explain these findings. A mid-term clinical study showed that focal osteolysis is more common with rough than polished stems that differed in no aspect other than their surface finish. Further studies established that focal osteolysis is probably always associated with defects in the cement mantle. These defects occur anteriorly at the mid-stem of the prosthesis and posteriorly at the component tip. The distribution of focal osteolysis and its strong association with cement mantle defects suggests the importance of the stemcement interface as a pathway for fluid and debris to reach the distal femur. However, at 15-25 years, osteolysis rarely develops with the polished Exeter stem even in the presence of confirmed defects in the cement mantle, suggesting that the stem seals the stem-cement interface against fluid and debris. In an attempt to explain the clinical findings a series of bench top experiments were undertaken. These studies showed that the behaviour of fluid and dye at the stemcement interface was significantly influenced by component surface finish. Bonded and debonded stem-cement interfaces of rough stems provided an incomplete barrier to fluid movement along this interface. In contrast, polished stems both bonded and debonded were able to provide a seal at the stem-cement interface. The seal at this interface was improved with component subsidence in the presence of rotational stability. It is believed that this thesis provides a rationale explanation for why focal osteolysis rarely develops around the Exeter stem in clinical practice. It also explains how, where, and why osteolysis develops around certain designs of cemented femoral components used in total hip replacement.