OBJECTIVE Defects of the anterior mitral leaflet (AML), including ruptured chordae, are often regarded as difficult or even impossible to repair. Chordal replacement may also be an option in extensive disease of the posterior mitral leaflet (PML). It has not yet been clearly defined whether the repair of either mitral leaflet using chordal-replacement techniques is as safe as the standard repair of the mitral valve (MV) including quadrangular resection and ring reduction alone. METHODS Between October 1995 and June 1999, 160 patients underwent MV repair for mitral regurgitation (MR) in our institution. Chordal replacement with polytetrafluoroethylene (PTFE) sutures for elongated or ruptured chordae was performed in 72 (45%) patients. These patients were divided into two groups according to the location of the MV lesions: 48 patients with prolapse of the anterior or both leaflets (AML group) received an average of 2.2+/-1. 1 PTFE sutures for repair; in 24 patients with isolated PML defects (PML group), we used an average of 1.5+/-0.8 PTFE sutures. No prosthetic annuloplasty rings were used. Dilatation of the posterior mitral ring was corrected by PTFE suture annuloplasty. The remaining 88 patients underwent a standard mitral repair without chordal replacement. There were no statistically significant (NS) differences between the two groups (AML/PML) regarding age (59/62 years, P=0.49), left ventricular (LV) ejection fraction (64/66%, P=0. 6) and preoperative NYHA class (2.9/2.9, P=0.36). Postoperatively, all patients were followed by serial transthoracic echocardiography at 1 week and after 3, 6, 12 and 24 months by the same investigator. RESULTS In-hospital mortality was 4.2% (2/48) in the AML group and 0% (0/24) in the PML group (P=0.55). Three of the AML patients (6. 3%) and one PML patient (4.2%) underwent reoperation for recurrent MR (P=1.0). The 1- and 2-year freedom from MV reoperation was 95. 1+/-3.4 and 92.6+/-4.2% in the AML group versus 95.0+/-4.9 and 95. 0+/-4.9% (P=0.67). The 1- and 2-year freedom from residual or recurrent MR grade 2 or higher was 97.6+/-2.4 and 94.9+/-3.5% (AML) versus 95.8+/-4.0 and 95.8+/-4.0% (PML) (P=0.97). CONCLUSIONS We were unable to find statistically significant differences concerning mortality, freedom from recurrent MR and MV reoperation between the AML and PML groups. Extensive prolapse or chordal pathology of the anterior and PML can be corrected by chordal replacement. Using these techniques, stable repair can be achieved in more than 90% of patients at mid-term follow-up. Long-term observations are necessary to confirm the durability of this type of MV repair.