Background: Physical activity (PA) has many beneficial effects, including improved fitness, strength, quality of life, and decreased insulin resistance. Several small studies suggest that HIV-infected adults report inadequate PA. We examined the relationship between HIV status, PA and insulin resistance among men in the Multicenter AIDS Cohort Study (MACS). Methodology: The International Physical Activity Questionnaire (IPAQ short form) was self-administered during MACS visits 53 and 54 (April 1, 2010 to March 31, 2011); we used data from the first administration that could be scored. Per IPAQ protocol, a metabolic equivalents (METs) total score (continuous variable) was calculated by adding MET-minutes/week for all activity performed and a categorical score (low, moderate, and high) was also generated. Concurrent homeostatic model assessment insulin resistance (HOMA-IR) (mmol/L) was calculated as: (fasting glucose x fasting insulin)/22.5. Additional covariates included age, race, BMI, education, smoking, HCV status, history of AIDS, ART, CD4, and VL. Adjusted quantile regression was used to investigate the effect of HIV status on PA (continuous). Adjusted linear regression was used to investigate the effect of PA (categorical) on logtransformed HOMA-IR by HIV status. Results: 1429 men (661 HIV-infected men, 768 HIV-uninfected men) were included in the analysis. HIV-infected status was significantly associated with higher total MET-minutes/week among men reporting the highest activity levels (75th percentile and up). However, the proportion of men reporting low (24% in HIV-infected vs. 23% in HIV-uninfected), moderate (27% vs. 28%), and high (49% vs. 49%) activity was similar by HIV status. Both HIV-infected status and low PA (vs. high PA) were associated with more insulin resistance (p<0.0001 and p=0.0007, respectively). Among men with low PA, HIVinfected men had higher HOMA-IR than their HIV-uninfected counterparts with similar activity levels (p<0.0001). However, this association was not seen among men in the moderate and high PA groups. Conclusions: In all men, lower PA was associated with more insulin resistance. However, HIV-infected men with the lowest levels of PA were more insulin resistant then HIV-uninfected men with a similarly low PA. This suggests that low PA may affect health outcomes more in HIV-infected persons compared to HIV-uninfected persons and that PA may need to be increased to a greater extent among HIV-infected persons to achieve the same metabolic benefit.