Associations between cognitive performance and cortisol have variously been reported for measures of both cortisol level and change, and for some domains of cognitive functioning more than others. In this study, associations between cortisol secretion measures and cognitive performance were examined in 50 healthy older people (mean age 74years; 34F /16M). Participants provided 16 accurately timed saliva samples over 2 consecutive days to determine diurnal profiles of cortisol secretion. Overall cognitive performance (OCP) was measured as the principal component of a comprehensive battery of cognitive tests. Across a 30year age range, there was a strong inverse correlation between age and OCP. Age and poorer OCP were also associated with an attenuated cortisol awakening response (CAR), defined as the rise from 0-30 minutes after awakening, and a subsequent less steep fall in cortisol level over the rest of the day. Partialling analyses, suggested that the correlation between fall in cortisol over the day and OCP was independent of age. Both older age and less cortisol change were particularly related to poorer performance on tests of declarative memory and executive functioning. Our conclusions are that during the short post-awakening period, an exception exists to the generally pertaining association between higher levels of cortisol and poorer cognitive performance. Consequentially dynamic measures reflecting the rise (CAR) and fall from the post-awakening peak may be particularly salient in helping to explain links between cortisol and cognitive performance. Finally our pattern of results across different cognitive tests suggests an association between cortisol and those domains of cognitive functioning which depend crucially on the integrity of the hippocampus and pre-frontal cortex.