Heart valve replacement with a bileaflet mechanical valve is a well-established procedure. However, the long-term results of valve replacement using the bileaflet mechanical valve remain unclear, especially for follow-up periods over 30 years. Additionally, it is important to identify predictors of long-term mortality and valve-related events. We performed a retrospective cohort analysis of 2727 patients (mean ± standard deviation age, 52.8 ± 1.6 years) who underwent valve replacement with a St. Jude Medical valve at our institute from 1978 to 2012. Data were collected using a questionnaire and chart review or physician contact. The cohort included 950 aortic valve replacements (AVRs), 1255 mitral valve replacements (MVRs) and 522 double valve replacements (DVRs). Follow-up was 91% complete, and the analysis included a total of 39 187 patient-years. Operative mortality rates were 2.3% for AVR, 2.2% for MVR and 3.6% for DVR. The 30-year survival rate (actuarial method) was 38.0% (AVR, 44.5%; MVR, 34.9%; and DVR, 37.5%). The 30-year rates of freedom from valve-related mortality, thromboembolic events and bleeding events were 86.3% (AVR, 88.6%; MVR, 85.4%; and DVR, 84.3%), 83.5% (AVR, 89.8%; MVR, 80.0%; and DVR, 81.4%) and 91.5% (AVR, 94.4%; MVR, 90.1%; and DVR, 90.2%), respectively. The incidence rates of valve-related morbidity, thromboembolic events and bleeding events were significantly higher among patients with MVR and DVR than among those with AVR. Significant risk factors for late death and other late events included male sex, age >65 years and atrial fibrillation. Low late mortality and a low incidence of valve-related events can be achieved for at least 30 years using mechanical bileaflet valve replacement. Persistent atrial fibrillation is a significant risk factor for morbidity and mortality.