Objectives: To assess the impact of a combination HIV prevention intervention including universal testing and treatment (UTT) on HIV stigma among people living with HIV, and among community members and health workers not living with HIV. Design: This HIV stigma study was nested in the HPTN 071 (PopART) trial, a three-arm cluster randomised trial conducted between 2013 and 2018 in 21 urban/peri-urban communities (12 in Zambia and nine in South Africa). Methods: Using an adjusted two-stage cluster-level analysis, controlling for baseline imbalances, we compared multiple domains of stigma between the trial arms at 36 months. Different domains of stigma were measured among three cohorts recruited across all study communities: 4178 randomly sampled adults aged 18-44 who were living with HIV, and 3487 randomly sampled adults and 1224 health workers who did not self-report living with HIV. Results: Prevalence of any stigma reported by people living with HIV at 36 months was 20.2% in arm A, 26.1% in arm B, and 19.1% in arm C (adjusted prevalence ratio, A vs. C 1.01 95% CI 0.49-2.08, B vs. C 1.34 95% CI 0.65-2.75). There were no significant differences between arms in any other measures of stigma across all three cohorts. All measures of stigma reduced over time (0.2--4.1% reduction between rounds) with most reductions statistically significant. Conclusion: We found little evidence that UTT either increased or decreased HIV stigma measured among people living with HIV, or among community members or health workers not living with HIV. Stigma reduced over time, but slowly.