An experimental model was designed to investigate the role of intramedullary pressure on cardiopulmonary function and pulmonary pathology during arthroplasty using cemented and non-cemented components. Twenty-four dogs were divided randomly into three groups: a group that received a non-cemented implant in which low intramedullary pressure was generated, a group that received a cemented implant, and one that received bone wax and an implant; high intramedullary pressures were generated in the latter two groups. Bone wax was used to generate high intramedullary pressures without the use of bone cement. In the group with the non-cemented implant, few pulmonary microemboli and no significant cardiorespiratory changes were found. In the groups that received bone wax and an implant or the cemented implant, there were many pulmonary microemboli and significant cardiorespiratory changes, including decreased arterial oxygen tension, increased pulmonary arterial pressure, and increased intrapulmonary shunt fraction. There was no evidence that methylmethacrylate monomer was responsible for the cardiorespiratory changes in the group with the cemented implant.