Mitral valvuloplasty has become an increasingly accepted alternative to valve replacement, although the strong learning curve attached to the procedure has deterred many surgeons from adopting it. Since it was developed in the late 1960s by Carpentier, comprehensive valvuloplasty has undergone modifications and additions dictated by the experiences thus acquired, which have contributed to make it a more reproducible and predictable procedure. Mitral valve disease, especially that of rheumatic origin, has a multifactorial pathogenesis, as all the components of the valve apparatus are usually involved. Consequently, valvuloplasty requires a combination of multiple techniques, each one directed at correcting each affected component. The excellent median-term results observed by some of the more experienced groups have contributed to change the indications for surgery, and patients are being referred earlier and in a lower functional class. Hence, in mitral valve prolapse surgery is currently offered to asymptomatic patients with significant regurgitation, before dysfunction of the myocardium and dilatation of the cardiac chambers occur. Valve repair has been possible and successful in virtually all patients with this pathological condition. On the other hand, surgeons are becoming more aggressive towards ischaemic mitral regurgitation of mild or moderate severity. More recently, valvuloplasty has also been performed in some patients with infective endocarditis, with good results. By contrast, the long-term results in rheumatic disease appear favourable, especially in children and young patients with acute carditis. Nevertheless, they have still been proven better than those of valve replacement in these population groups.(ABSTRACT TRUNCATED AT 250 WORDS)