The aim of this long-term case series was to assess the development/prevalence of non-carious cervical lesions (NCCLs) at sites that have and have not been treated with gingival augmentation following free gingival graft (FGG).Fifty-two patients had at least one test and one control site: 1) test site showing absence of attached gingiva (AG) associated with gingival recession (GR) treated with FGG; and 2) contralateral site with or without AG. Patient/tooth/site-associated variables were recorded for each tooth/site at baseline (T0), 12 months after surgery (T1), during the follow-up period (T2) (15 to 20 years), and at the end of the follow-up period (T3) over 25 to 30 years. Mixed-effects logistic regression was used throughout the study.Forty-nine patients/130 sites were available for analysis at T2 whereas 44 patients/120 sites at T3. Twenty-two NCCLs >0.5 mm were restored in the test sites and in 35 in the untreated sites. The development of NCCL over time appeared associated with sites with attached KT <2 mm (i.e., odds ratio [OR]: 3.80 [P = 0.045] and 3.47 [P = 0.046], 15- to 20- and 20- to 30-year follow-ups, respectively), as well as to teeth presenting a thin/non-modified periodontal phenotype (i.e., OR: 3.53 [P = 0.037] and 5.51 [P = 0.008], 15- to 20- and 20- to 30-year follow-ups, respectively).Periodontal phenotype modification achieved by FGG may prevent the development/progression of NCCL. Evidence suggests that the thickness and width of the AG had a direct influence on the need of restoring these lesions during the 25- to 30-year observation period.