Several studies have suggested that aerobic physical activity is safe and beneficial for HIV-infected adults. However, there is information lacking regarding whether HIV-infected patients practice physical activity and to what extent. Therefore, the aim of this systematic review was to determine the prevalence of physical activity, sedentary lifestyle or lack of physical activity in non-experimental conditions performed by HIV-infected subjects. The electronic search was conducted using Medline and EMBASE bibliographic databases and the platforms of Bireme, Ovid, Science Direct, High Wire and SCIELO from January 1990 to July 2011. Original observational studies were included. Of the 2,838 articles found, 48 met the inclusion criteria. Following data extraction and after reading the manuscripts, 24 were selected for systematic review. Of the 24 studies, most were cross-sectional studies. The average quality score using the modified Newcastle-Ottawa scale was 2.8±1.5. The diversity of methods used to assess physical activity precluded the calculated summary estimate of prevalence. The percentage of sedentary lifestyle was determined in 13 articles which conducted studies on HIV-infected individuals. The percentage of sedentary lifestyle or physical inactivity ranged from 19% to 73%, with the level determined by different methods. In conclusion, there are few well-designed studies with adequate sample size to represent the population of HIV-infected individuals. A pooled estimate could not be calculated due to the differences in physical activity measurements and definitions of physically active and non-active HIV-infected individuals.