We examined the association between CD4 cell count and adherence in a cohort of Ugandans initiating antiretrovirals (ARVs). Outcomes were (a) adherence <90%; (b) any treatment interruptions > 72 hours; (c) number of treatment interruptions; and (d) HIV-RNA >400 copies/mL. We fit regression models to estimate associations with our exposure of interest, baseline CD4 cell count ≥ 250 cells/μL (n = 60) vs <250 cells/μL (n = 413). CD4 cell count ≥250 cells/μL was independently associated with increased odds and number of treatment interruptions and increased odds of persistent viremia. Interventions to support adherence in patients with higher CD4 cell counts should be considered as drug availability to this population increases.