BACKGROUND: This study examined the construct validity and reliability of the new 2-item Stanford Brief Activity Survey (SBAS). METHODS: Secondary analysis was conducted using data collected from the healthy older controls (n = 1023) enrolled in the Atherosclerotic Disease Vascular Function and Genetic Epidemiology (ADVANCE) study. Construct validity was examined by regression analyses to evaluate significant trends (P < or = .05) across the SBAS activity categories for the selected psychological health factors measured at baseline and year 2, adjusted for gender, ethnicity and education level. Test-retest reliability was performed using Spearman's rank correlation. RESULTS: At baseline, subjects were 66 +/- 2.8 years old, 38% female, 77% married, 61% retired, 24% college graduate, and 68% Caucasian. At baseline, lower self-reported stress, anxiety, depression, and cynical distrust, and higher self-reported mental and physical well-being were significantly associated with higher levels of physical activity (p trend < or = 0.01). These associations held at year 2. The test-retest reliability of the SBAS was statistically significant (r(s)= 0.62, P < .001). CONCLUSION: These results provide evidence of the construct validity and reliability of the SBAS in older adults. We also found a strong dose-response relationship between regular physical activity and psychological health in older adults, independent of gender, education level and ethnicity.