Abstract Background In the general population, it is established that adipose tissue depots pose various risks for cardiometabolic diseases. The interaction between obesity, HIV and anti-retroviral treatment promotes even greater risk for persons with HIV (PWH). As obesity is a heterogeneous condition, determining the specific obesity phenotypes present and their characteristics is critical to personalize care in PWH. Methods Visceral (VO), sarcopenic (SO), myosteatotic (MO), hepatosteatotic (HO) and metabolically healthy (MHO) obesity phenotypes were determined using pre-established cut-points after segmentation of CT scans at the L3 vertebra. Multivariable linear regression modeling included anthropometrics, clinical biomarkers, and inflammatory factors while controlling for age, sex, race, and BMI. Results Of 187 PWH, 86% were male, age was 51.2 ± 12.3 years, and BMI was 32.6 ± 6.3 kg/m2. 59% had visceral obesity (VO), 11% sarcopenic obesity (SO), 25% myosteatotic obesity (MO), 9% hepatosteatotic obesity (HO), and 32% metabolically healthy obesity (MHO). The strongest predictor of VO was elevated TG:HDL ratio. Increased subcutaneous fat, waist circumference, and HDL-cholesterol were predictors of SO. Diabetes status and elevated IL-6, waist circumference, and HDL-cholesterol predicted MO. Increased CD4+ count and decreased VAT:SAT ratio predicted HO; however, only accounting for 28% of HO variability. MHO participants averaged 10 years younger, had higher HDL, lower TG:HDL ratio, and lower CD4+ counts. Conclusions These findings show discrete obesity phenotypes are highly prevalent in PWH and convey specific risk factors that measuring BMI alone does not capture. These clinically relevant findings can be used in risk stratification and optimizing personalized treatment regimens. This study is registered with clinicaltrials.gov identifier NCT04451980