Background: Inconsistent findings exist regarding correlates of physical activity (PA) in the literature. Leisure-time physical activity among U.S. adults has declined for the last decade. Purpose: This article examines differences in vigorous-intensity and moderate-intensity physical activity by gender, race/ethnicity, age, education, and income among a representative sample of U.S. adults. Methods: A total of 1,000 adults participated in a random-digit telephone survey in 2005 (62% response rate and 82% cooperation rate). Non-Hispanic Blacks and Hispanics were oversampled to produce reliable estimates for these groups. Standard errors and odds ratios were calculated utilizing SUDAAN, reflecting differential probabilities of selection. Results: After adjusting for differences in existence of a serious health problem, body mass index, marital status, employment status, and smoking status in the multivariable logistic models, none of the five demographic variables except race/ethnicity were predictive of meeting the moderate PA guideline, whereas each of the five variables was predictive of meeting the vigorous PA guideline. Discussion: Because most U.S. adults fail to meet physical activity guidelines, health practitioners must be vigilant in helping people adopt and adhere to active lifestyles. Translation to Health Education Practice: The results of this study support different approaches for different intensity levels of PA. Given the lack of evidence about the efficacy of moderate-intensity PA on reducing body weight, health educators must be cautious against providing expectation that people could lose substantial weight by engaging in moderate PA. (Contains 2 tables.)