The MitraClip technique has been increasingly used for correction of mitral valve regurgitation in patients in whom surgical mitral repair is considered contraindicated or very risky, but off label use occurs often. Failure of the procedure, translated into moderate to severe rates of residual or recurrent mitral regurgitation, is observed in up to one-third of the patients, and surgery has been used to correct it in a number of cases, in what can be called an “operation for the inoperable”. That is precisely the subtitle of a paper published in this issue of the JOCS by Gerfen and colleagues, who analyse their institutional experience with a series of 17 patients. In this Editorial, I comment on this series and the possible reasons for failure of the MitraClip, and on the indications for reintervention and its constraints, which I hope can contribute to the discussion about “further exploration and refinement of patient selection criteria and identify predictors for MitraClip failure”, as the authors suggest.