This epidemiological study aimed at determining the prevalence of mild cognitive impairment (MCI) in 60- to 64-year-old individuals using different diagnostic criteria. Community dwelling individuals (n = 2,551) in the age range of 60-64 years were recruited randomly through the electoral roll. They were screened using the MMSE and a short cognitive battery, and those who screened positive underwent detailed medical and cognitive assessments. Extant MCI-related diagnoses were established by consensus. Predictive regression models on the sub-sample were used to determine population prevalence for the diagnoses. Of the 224 subjects who screened positive for MCI, 112 underwent a detailed assessment and 74% met the criteria for at least one recognised diagnosis of mild cognitive deficit (MCI and related diagnoses). By predictive regression modelling, the prevalence of any MCI diagnosis was 13.7% (95% CI 9.1-30.2) in the population of 60- to 64-year-olds. The estimated prevalence rates for specific diagnoses were: MCI 3.7%, ageing-associated cognitive decline 3.1%, Clinical Dementia Rating score (0.5) 2.8%, age-associated memory impairment 1%, other cognitive disorders 0.9%, and mild neurocognitive disorder 0.6%. Agreement on 'caseness' between various proposed diagnoses was at best fair and generally poor. Memory and other cognitive problems not meeting the threshold for dementia are relatively common in 60- to 64-year-old individuals living in the community. The prevalence rate varies up to six-fold according to the diagnostic criteria applied, with limited overlap between diagnoses. There is an urgent need for standardization of the criteria.