The primary purpose of this study was to evaluate the long term performance of several prosthetic cardiac valves and the influence on outcome of different surgeons by following 328 patients for up to 10 years after mitral valve replacement surgery in a tertiary care centre. This type of follow-up study is faced with severe methodological difficulties, as mortality and complications after surgery can be influenced not only by preoperative clinical status but also by the competence of the operating surgeons, the type of prosthesis used, and system-related variables such as the overall quality of nursing care. The primary statistical models used were life-table analysis, logistic regression and Cox's proportional hazard regression, employing prosthesis type, surgeon and other patient- or system-specific variable as covariates. The relative performance of valve prostheses and surgeons were assessed for short and long term survival. Biologic and Omniscience valves outperformed others for short and long term survival, while Lillehei-Kaster valves performed well for short term but not long term survival. The results also suggest that the performance levels of surgeons do vary, and the choice of surgeon could be a significant risk factor in a given medical centre.