Background: Time updated VL levels and CD4 counts are routinely used to monitor HIV positive adults but fail to capture cumulative HIV exposure. Viremia copy years (VCY) is such a measure and has been shown to be predictive of AIDS/death, although it is unclear of its use in deciding when to start ART. We aimed to assess the impact of initiating vs. deferring ART on risk of AIDS/death by levels of VCY both independent of and within two CD4 strata, <350 and ≥350 cells/mm3. Methodology: Using CASCADE data on HIV seroconverters, we created a series of nested cohorts corresponding to consecutive months starting month 5 post seroconversion (SC) for individuals ≥16 years at SC after 1997, had ≥1 VL measured 4-12 months post SC, were ART naive and AIDS free prior to end of baseline month. Time to AIDS/death was compared in those initiating vs. deferring ART in the baseline month using Cox models adjusted for time independent factors: country, sex, risk group, SC year; and time dependent factors prior to the baseline month: age, time since last VL, and current CD4, VCY and VL (excluding the first 4 months post SC) and mean number of previous CD4/VL measurements/year. Robust variance was used to adjust for individuals contributing ≥1 baseline visit. We repeated analyses using CD4 strata <500 and ≥500 cells/mm3. Results: Of 6497 individuals contributing ≥1 baseline month, 3089 (48%) initiated ART and 293 (5%) acquired AIDS/died. Median (IQR) CD4 at ART initiation was 322 (249, 419) cells/mm3. Pooling CD4 strata, hazard ratios (HR) of AIDS/death associated with initiating vs. deferring ART reduced as VCY increased, suggesting a stronger benefit of ART in those with higher VCY. Trends by VCY were different when stratifying by CD4; in patients with CD4<350 cells/mm3, there was an overall reduction in the HR of AIDS/death in all VCY groups (all HR < 1) with no evidence that this benefit varied by VCY (p=0.78). At CD4 ≥ 350 there was a trend for increasing benefit of initiation with increasing VCY, with the largest benefit seen in the VCY ≥ 100,000 c/mL group (HR, 95% CI= 0.56, 0.35-0.90); no evidence of benefit if VCY was < 20,000 c/mL. Results were qualitatively similar for CD4 strata ≥ 500 cells/mm3. Conclusions: Although we cannot rule out the possibility of unmeasured confounding, it appears that limiting the cumulative HIV burden to < 100,000 VCY in individuals with CD4 ≥ 350 cells/mm3 through ART reduces the hazard of clinical outcomes by 44%.