ABSTRACT Timely initiation of antiretroviral therapy (ART) remains a major challenge in the effort to treat children living with HIV (“CLH”) and little is known regarding the dynamics of immune normalization following ART in CLH with varying times to and durations of ART. Here, we leveraged two cohorts of virally-suppressed CLH from Nairobi, Kenya to examine differences in the peripheral immune systems between two cohorts of age-matched children (to control for immune changes with age): one group which initiated ART during early HIV infection and had been on ART for 5-6 years at evaluation (early, long-term treated; “ELT” cohort), and one group which initiated ART later and had been on ART for approximately 9 months at evaluation (delayed, short-term treated; “DST” cohort). We profiled PBMC and purified NK cells from these two cohorts by mass cytometry time-of-flight (CyTOF). Although both groups of CLH had undetectable viral RNA load at evaluation, there were marked differences in both immune composition and immune phenotype between the ELT cohort and the DST cohort. DST donors had reduced CD4 T cell percentages, decreased naive to effector memory T cell ratios, and markedly higher expression of stress-induced markers. Conversely, ELT donors had higher naive to effector memory T cell ratios, low expression of stress-induced markers, and increased expression of markers associated with an effective antiviral response and resolution of inflammation. Collectively, our results demonstrate key differences in the immune systems of virally-suppressed CLH with different ages at ART initiation and durations of treatment and provide further rationale for emphasizing early onset of ART. AUTHOR SUMMARY Many children living with HIV lack access to both antiviral treatments and testing for HIV infection and are therefore unable to initiate treatment in a timely manner. When children do begin treatment, their immune systems take time to recover from the uncontrolled HIV infection. In this study, we examine how the immune systems of children living with HIV normalize after treatment onset by looking at two groups of children whose HIV is well-controlled by treatment and who therefore don’t have virus replicating in their blood. One group started treatment within the first year of life and has been on treatment for 5-6 years, while the other began treatment after the first year and has been treated for around 9 months. Although both of these groups are virally-suppressed, we found significant differences in their immune profiles, with the children who had delayed and short-term treatment showing signs of inflammation and immune dysfunction. Collectively, our study helps us understand how variation in the timing and duration of ART treatment impacts the immune system in children with viral suppression and therefore provides clinicians with additional knowledge that can inform the care of children living with HIV, improving their health and quality of life.