Social perception is an important aspect of social communication. Accurate recognition of emotion is crucial to understanding how others are feeling, which in turn is important in developing and maintaining meaningful relationships. Difficulties in social perception may lead to feelings of loneliness and isolation, which affects quality of life and wellbeing. Social support is particularly important, as people get older as it helps foster wellbeing. The process of ageing is related to a gradual and slow decline in cognitive function. This decline is seen to hasten with the onset of dementia. The risk of dementia is greater in people with mild cognitive impairment (MCI). There has been limited research examining social perception in people with MCI and existing research has largely used static visual tasks, such as photographs and stories for assessment of social perception. Further, those studies have predominantly focussed on people with MCI that are more likely to progress to Alzheimer’s disease, i.e., amnestic MCI. The present study aimed to assess emotion recognition and complex social perception tasks using an ecologically valid tool - The Awareness of Social Inference Test (TASIT) - in all subtypes of MCI. The overall aims were to examine social perception skills in people with MCI and to examine the relationship between social perception and neuropsychological functioning. Ninety-six healthy participants (females = 69, M age = 62.74, SD = 8.1) above the age of 50 years were recruited from the community, and 21 participants (females =10, M age = 72.24, SD = 10.4) with a diagnosis of MCI were recruited from the local District Health Boards (DHBs). These participants were administered a battery of neuropsychological tests (memory, language, executive functioning and visuospatial functioning domains), the TASIT, social activity and network measures, and informant measures were also collected. Initially, analyses were conducted to determine which of four MCI classification systems (liberal, comprehensive, conventional and conservative) was the most appropriate for use in the current study. This analysis revealed that the conventional criteria (1.5 SD below mean in one measure of a domain) accurately classified all the participants with a diagnosis of MCI, and 49% of the community sample as meeting the criteria for MCI. Overall, 68 participants (58%) met the criteria according to the conventional criteria and 49 participants (42%) were in the non-MCI group. Analyses revealed no significant differences between the MCI and non-MCI groups on any of the TASIT subtests. In addition, there were no significant differences on TASIT scores between those with multiple domain MCI compared to those with single domain MCI and participants with no MCI. There was no significant association between meeting criteria for MCI and the proportion of participants meeting the cut-off score for moderate deficits on the TASIT. When examining the sub-tests of the TASIT, there was a greater proportion of participants who had deficits on the recognition of emotion (76%) compared to the higher order social perception tests (41%). Finally, all participants obtained significantly higher scores on the recognition of negative emotions, suggesting this was easier than the recognition of positive emotions. A series of correlations between the TASIT sub-test scores and the neuropsychological domains scores were conducted. The analyses revealed no significant correlations between any of these measures. There were no significant differences between groups on their informant ratings. Overall, the findings suggest that MCI is not associated with a decline in social perception, as assessed by the TASIT. Furthermore, the results of this study support the suggestion that social perception may be viewed as a separate domain of functioning, which could be assessed separately as part of a comprehensive neuropsychological assessment. This may assist in the accurate diagnosis of non-amnestic types of MCI, which are more likely to progress to a non-Alzheimer’s dementia.