Objective: This retrospective study aimed to evaluate the long-term response of molars affected with severe periodontitis to periodontal treatment, to analyze the differences in response between molars with and without furcation involvement (FI) and to reevaluate the usefulness of the existing classification of FI in determining the prognosis. Method and Materials: A total of 402 patients from a single private practice were included. The observation period was 27 years (mean 16.5 years). The average frequency of the supportive periodontal therapy (SPT) visits was 1.76 ± 0.57 per year (median 1.95/ year). Inclusion criteria were: at least 10 years of periodontal follow-up, at least one tooth with probing depth ≥ 6 mm, level 4 of the Dutch Periodontal Screening Index (DPSI). Results: Of the 2,559 molars present at the initial examination, 125 were extracted immediately. Degrees III or II FI were found in 37.2% molars, while 62.8% exhibited degree I or had no FI. The performed periodontal treatments were: nonsurgical therapy of scaling and root planing (77.6%), eventually repeated scaling and root planing (11.6%), access flap (7.8%), tunnellization (0.2%), and root resection (2.8%). The survival rate was 83.9% (77.5% molars with FI; 87.8% without FI). The presence or absence of FI did not exhibit a significant effect in any of the treatments provided. Conclusion: Simple treatments can successfully be applied to treat molars affected with severe periodontitis, even with FI, yielding good long-term survival rates. The classification of FI was not of great use in determining the prognosis except in case of degree III FI. Keywords: furcation defect, long-term care, molar, retrospective study, risk factor