There are conflicting data on whether variations of physiologic cortisol levels associated with cardiovascular risk. We hypothesize that prior discordant findings are related to problems associated with varying sample size, techniques for assessing cardiovascular risk and failure to adequately account for environmental factors. To address these issues, we utilized a large sample size, selected the Framingham risk score to compute cardiovascular risk and performed the study in a highly controlled setting. We had two main objectives: determine whether higher, yet physiologic, cortisol levels associated with increased cardiovascular risk and determine whether caveolin-1 (rs926198) risk allele carriers associated with increased cardiovascular risk. This was a cross-sectional study of 574 non-diabetic individuals who completed a common protocol. Data collection included fasting blood samples, blood pressure measurements and a 24-h urine-free cortisol collection. Five hundred seventeen of these participants also completed caveolin-1 genotyping. Subjects were classified as belonging to either the low-mode or high-mode urine-free cortisol groups, based on the bimodal distribution of urine-free cortisol. In multivariate analysis, Framingham risk score was statistically higher in the high-mode cortisol group (10.22 (mean) ± 0.43 (s.e.m.)) compared to the low-mode cortisol group (7.73 ± 0.34), P < 0.001. Framingham risk score was also statistically higher in the caveolin-1 risk allele carriers (8.91 ± 0.37) compared to caveolin-1 non-risk allele carriers (7.59 ± 0.48), P = 0.034. Overall, the estimated effect on Framingham risk score of carrying the caveolin-1 risk allele was 1.33 ± 0.61, P = 0.029. Both urinary cortisol and caveolin-1 risk allele status are independent predictors of Framingham risk score.