The objective of this study was to assess the use of HIV-oral lesions (OLs) as markers of virologic failure (VF) in response to antiretroviral therapy (ART). Concurrent virologic status was compared between 744 individuals with and without OL (). Time to VF between OL groups was compared (Kaplan-Meier), and baseline factors associated with VF were determined (Cox Hazard models). Sensitivity, specificity, positive predictive value, and negative predictive value were also computed. At baseline and 12 months, individuals with OL were more likely to have viral load copies/ml compared to patients without OL (). Time to VF between patients with baseline OL (mean: 17 months, 95% CI 16–18) and patients without baseline OL (mean: 19 months, 95% CI: 18–20) was statistically different. Patients who were African-American (HR 1.356; 95% CI: 1.045–0.759), ART-experienced (HR 2.298; 95% CI: 1.743–3.030), had mental disorders (HR 1.410; 95% CI: 1.078–1.843), and had high baseline viral load (HR 2.82; 95% CI: 1.661–3.137), were more likely to have VF after the first six months of ART. OL had a moderate positive predictive value for concurrent VF at 6 months (45.5%) and 18 months (33.3%), but a strong positive predictive value at 12 months (80.0%) and 24 months (100.0%). Findings of this study suggest that OLs could be poor predictors of VF in HIV-infected patients on ART therapy.