To analyze the clinical results of repeat Descemet membrane endothelial keratoplasty (re-DMEK) for failed primary DMEK graft at a referral center for keratoplasty in Spain. Retrospective, interventional, comparative case series. From a single-center, single-surgeon series of 189 consecutive DMEK surgeries, 14 (7.41%) were re-DMEK eyes. The primary outcome was best-corrected visual acuity change (ΔBCVA, logarithm of minimal angle of resolution [logMAR]) from baseline (before first DMEK) to last follow-up. Secondary outcomes were ΔBCVA from baseline at 3, 6, and 12 months postoperatively, endothelial cell loss (%ECL), rebubbling rate, and re-DMEK graft failure. Outcomes were compared with an age-matched control group of 18 successful primary DMEK eyes. After re-DMEK (median follow-up time 14.5 [42.5] months), mean BCVA improved from 0.55 (0.42) logMAR (Snellen 20/71 [20/53]) at baseline to 0.09 (0.26) logMAR (Snellen 20/25 [20/36]; P = .037). ΔBCVA from baseline was statistically significant at months 3 (P = .028), 6 (P = .023), and 12 (P = .012), and ΔBCVA was significant observed between months 6 and 12 (P = .028). BCVA differences between patient groups were statistically nonsignificant at 3 (P = .397), 6 (P = .468), and 12 months (P = .647). Mean %ECL in re-DMEK eyes with follow-up ≥12 months was 48.2 (15.1%), and the rebubbling rate was 28.6%; differences between groups were statistically nonsignificant for both variables (P = .580 and P = .669, respectively). Three re-DMEK eyes developed graft failure, all achieving final BCVA ≤0.30 logMAR (Snellen ≥20/40) after tertiary keratoplasty. Repeat DMEK produces significant, continuous visual improvement after failed primary DMEK. Although visual outcomes and %ECL were comparable to primary DMEK, there was a relatively high rate of graft failure after re-DMEK.