Success rates for both periodontal and implant therapy are often dependent on site and tooth type. For periodontally involved mandibular molars, the decision to hemisect or to extract and place an implant is often complicated. The purpose of the present study was to evaluate the outcomes of the aforementioned treatment modalities for mandibular molars in a private practice setting. A retrospective chart review was performed. In one group of patients (n = 32), 56 mandibular first or first and second molars were treated by hemisection (Group H). A second group (n = 28) received 36 implants in the mandible to replace periodontally involved first or first and second molars (Group I). All patients had been in maintenance for at least 4 years after treatment. The occurrence and timing of posttreatment complications were evaluated. Data were analyzed by parametric and nonparametric statistics, as indicated. The majority of hemisected teeth (68% of Group H) and implants (89% of Group I) remained free of complications for the entire observation period. Group H had a greater incidence of overall complications (P = .027) and nonsalvageable complications (P = .013) than Group I. For both groups, the percent CAL loss per year was greater for the teeth/implants that experienced complications than in the those that remained complication free (p < 0.015). Within the limitations of this study, the results indicated that, in periodontitis patients, hemisected mandibular molars were more prone to complications than implants.