Chronic stress detrimentally affects cognition but evidence from population-based studies is scarce and largely based on one-dimensional stress assessments. In this study, we aimed to investigate associations of subjective and psychological chronic stress measures with cognition in a population-based sample of adults aged 30-95 years from the Rhineland Study. Participants completed the Perceived Stress Scale (subjective measure) and a cognitive test battery (N = 1766). Hair cortisol concentration (physiological measure) was assessed by liquid chromatography tandem mass spectrometry in 1098 participants. Cross-sectional associations between the two measures of chronic stress and cognition were investigated using multivariable linear regression models. Subjective and physiological measures of chronic stress were not associated with each other (B = 0.005 [95 %CI = -0.005 - 0.015]). Participants with higher perceived stress and specifically lower perceived self-efficacy performed worse in all cognitive domains (effect sizes ranged from β = -0.129 [95 %CI = -0.177 - -0.080] to -0.054 [95 %CI = -0.099 - -0.009]; and from β = 0.052 [95 %CI = 0.005 - 0.098] to 0.120 [95 %CI = 0.072 - 0.167], respectively). Relationships between subjective chronic stress measures and executive functioning were stronger in men compared to women (interaction β = -0.144 [95 %CI = -0.221 - -0.067]). Relationships between perceived stress and working memory, and between perceived self-efficacy and executive functioning, processing speed, verbal episodic and working memory, increased with older age. Hair cortisol concentration was not associated with performance in any cognitive domain. Our results suggest that subjective and physiological measures capture different aspects of chronic stress in the general population.