OBJECTIVES Arthritis is the most common chronic illness in the US. Various studies have found association between arthritis and physical demands of work. This study determines the relationship between physical activity at work and prevalence of arthritis among working population. METHODS To distinguish between arthritic and non arthritic population we used question from the 2007 Behavioral Risk Factor Surveillance System (BRFSS) that indicates whether an individual was suffering from arthritis that had been diagnosed by physician. We also determined an individual level of physical activity at work through the question that classifies an individual level of physical activity at work into three levels. RESULTS The sample consists of 205,533 respondents out of which (46, 078) 22.42% had arthritis and (15, 9455) 77.58% did not have any type of arthritis. Results from the logistic regression showed significant relationship between prevalence of arthritis and gender, age, race, education, income, reported health status, BMI, health coverage and physical activity at work (α=0.05). There was inconsistent relationship among the level of physical activity and prevalence of arthritis as people having moderate physical activity had lesser odds (O.R.-0.965, 95% CI: 0.938-0.994) of suffering from arthritis i.e. one unit increase in moderate physical activity at work was associated with 3.5% decrease in the odds of suffering from arthritis. However, people having heavy physical activity at work had greater odds of (O.R.-1.265, 95% CI: 1.220-1.311) of suffering from arthritis indicating that one unit increase in heavy physical activity was associated with 27% increase in the odds of suffering from arthritis. CONCLUSIONS There is a need to investigate in greater detail the role of physical activity at work in conjunction with other factors on the prevalence of arthritis. This investigation can help in identifying people susceptible to develop arthritis along with the factors responsible for their illness.