BACKGROUND: Retrospective and prospective memory deteriorate with age and deteriorate further with the onset of dementia. As previous research has tended to use idiosyncratic and heterogeneous methodologies, it is not known if the deterioration in retrospective or prospective memory is equal or how such deterioration is related to insight into mnemonic performance. DESIGN: The present study used a mixed, cross-sectional design. It examined retrospective and prospective memory in healthy and cognitively impaired older adults using objective and subjective assessment. METHODOLOGY: Twenty healthy and twenty cognitively impaired older adults were assessed. The objective measures were standardised and had high ecological validity, to maximise the assessments’ generalisability to everyday mnemonic functioning. The subjective measure was a standardised assessment of subjective appraisal of retrospective and prospective memory. This subjective assessment was used to gain self-ratings from the healthy and cognitively impaired older adults and, in addition, to gain proxy-ratings from the cognitively impaired older adults’ partners/carers. Participants were also assessed using a measure of general cognitive functioning. RESULTS: Healthy older adults performed significantly better than cognitively impaired older adults on both the retrospective and prospective memory test, although performance in both groups was higher on the prospective memory test than on the retrospective memory test. Neither proxy-ratings, nor overall severity of impairment, were associated with subjective assessment of prospective or retrospective memory. In the healthy older adults, self-rating of memory was not associated with objective memory performance. In the cognitively impaired older adults, however, higher self-ratings of memory were associated with poorer performance on the prospective memory test. DISCUSSION: The results suggested that both healthy and cognitively impaired older adults perform better on prospective memory tasks than retrospective memory tasks, but both types of memory deteriorate with the onset of dementia. Subjective memory appraisal is not related to objective memory performance in healthy older adults, but is negatively associated with prospective memory performance in cognitively impaired older adults. Findings were discussed in relation to previous research, along with a review of the strengths and limitations of the study. Clinical implications and directions for future research are also suggested.