OBJECTIVE: We sought to evaluate factors that are associated with physical activity changes among adults within a 5-year period. METHODS: From May 2001 to December 2002 we randomly enrolled 1514 men and 1528 women without any evidence of cardiovascular or any other chronic disease. The sampling was stratified by the age-gender distribution of the greater area of Athens. Weekly energy expenditure assessed by considering frequency, duration, and intensity of sports-related physical activity. During 2006, the 5-year follow-up was performed in 1955 participants, which included, among others, current physical activity status. RESULTS: A total of 587 (61%) men and 673 (68%) women were classified as physically inactive at baseline, whereas 661 (69%) men and 728 (73%) women were classified as physically inactive at follow-up; thus, a 13% increase in physical inactivity rate was observed in men and a 7% in women during the follow-up period (p < 0.01). Multivariate analysis revealed that male gender, being a nonsmoker, healthy eating, better self-reported quality-of-life, and lower prevalence of hypercholesterolemia and incidence of cardiovascular disease were the characteristics of people that remained physically active; advanced age, anxiety and depression, overweight, and low quality-of-life were the baseline predictors of physical inactivity among initially active participants, which also had a greater incidence of cardiovascular disease. CONCLUSIONS: Gender, aging, psychological disorders, body mass, smoking, dietary habits, perceived health status, and quality-of-life were the most important discriminating factors of physical activity changes.