OBJECTIVES: To determine how muscle strength, power, mass, and density (i.e. quality) differ between children living with HIV (CWH) and those uninfected, and whether antiretroviral therapy (ART) regime is associated with muscle quality. DESIGN: A cross-sectional study in Harare, Zimbabwe. METHODS: The study recruited CWH aged 8-16 years, taking ART for at least 2 years, from HIV clinics, and HIV-uninfected children from local schools. Muscle outcomes comprised grip strength measured by hand-held Jamar dynamometer, lower limb power measured by standing long-jump distance, lean mass measured by dual-energy X-ray absorptiometry, and muscle density (reflecting intramuscular fat) by peripheral quantitative computed tomography. Linear regression calculated adjusted mean differences (aMD) by HIV status. RESULTS: Overall, 303 CWH and 306 without HIV, had mean (SD) age 12.5 (2.5) years, BMI 17.5 (2.8), with 50% girls. Height and fat mass were lower in CWH, mean differences (SE) 7.4 (1.1) cm and 2.7 (0.4)kgs, respectively. Male CWH had lower grip strength [aMD 2.5 (1.1-3.9) kg, P  < 0.001], long-jump distance [7.1 (1.8-12.5) cm, P  = 0.006], muscle density [0.58 (0.12-1.05) mg/cm 3 , P  = 0.018, but not lean mass 0.06 (-1.08 to 1.21) kg, P  = 0.891) versus boys without HIV; differences were consistent but smaller in girls. Mediation analysis suggested the negative effect of HIV on jumping power in boys was partially mediated by muscle density ( P  = 0.032). CWH taking tenofovir disoproxil fumarate (TDF) had lower muscle density [0.56 (0.00-1.13)mg/cm 3 , P  = 0.049] independent of fat mass, than CWH on other ART. CONCLUSION: Perinatally acquired HIV is associated, particularly in male individuals, with reduced upper and lower limb muscle function, not mass. Intra-muscular fat (poorer muscle quality) partially explained reductions in lower limb function. TDF is a novel risk factor for impaired muscle quality.