Objective: Cortisol levels increase with age and hypercortisolism is associated with muscle weakness. This study examines the relationship between cortisol, muscle mass and muscle strength in community-dwelling older persons and the role of genetic variations in the glucocorticoid receptor (GR). Design/patients: The study was conducted within the Longitudinal Ageing Study Amsterdam (LASA, 1992-ongoing), a cohort study in a population-based sample of older persons in the Netherlands. Data were used from 1196 and 1046 participants in the second (1995-1996) and fourth (2001-2002) cycle, respectively. Measurements: Total serum cortisol and free cortisol were measured in the mornings of the second cycle while salivary cortisol sampled early in the morning and late at night were measured in the fourth cycle. The GR gene polymorphisms (ER22/23EK, N363SS, 9β and BclI) were genotyped by Taqman. Appendicular skeletal muscle mass (ASMM) was measured using DXA in the second cycle and 3 years later (third cycle). Grip strength was assessed using a handgrip dynamometer in the second, third, fourth and fifth cycle. Results: A relationship was found between both morning and evening salivary cortisol, and loss of grip strength: participants in the highest quartile of cortisol concentration had a twofold higher risk of loss of grip strength than participants in the lowest quartile (P < 0.05). No relationships were found between serum cortisol (loss of) ASMM, and (loss of) grip strength. The ER22/23EK and N363S-polymorphisms modified the relationships between serum cortisol, ASMM and grip strength, respectively. Due to limited power, these relationships were not significant after stratification for the polymorphisms. Conclusion: High salivary cortisol is associated with a higher risk of loss of grip strength in older persons. GR genotypes modify the relationship between muscle mass and muscle strength. © 2008 The Authors.