OBJECTIVE: The present study used information from a field study conducted among 4489 civil servants (70% women) in Denmark in 2007. The purpose was to examine the association between sleep problems and salivary cortisol by using a cross-sectional design with repeated measures in a subsample three-month later. METHODS: Sleep problems during the past night and the past 4 weeks were assessed by a self-administered questionnaire on overall sleep quality, disturbed sleep, sleep length and awakening problems. Saliva samples were collected in a single day, using cotton tubes, 30min after awakening and again at 2000h. A subsample of 387 participants collected saliva samples three-month later at awakening, +20min and +40min after awakening and at 2000h. We adjusted for confounders related to sampling time, life style and personal characteristics, socioeconomic status and work aspects. RESULTS: Sleep problems during the past four weeks were associated with low morning and evening saliva cortisol concentrations: [-3.1% per score of disturbed sleep (p=.009); and -4.7% per score of awakening problems (p