BACKGROUND: The research criteria for subjective cognitive decline (SCD) exclude mild cognitive impairment (MCI), but do not stipulate the use of specific MCI criteria. This study compared different approaches to defining (i.e., excluding) MCI during the ascertainment of SCD, focusing on the impact on dementia incidence rates in SCD. METHODS: This cohort study utilized routine healthcare data collected in the Essex Memory Clinic from 1999 to 2023. Two different operationalizations of the SCD criteria were used to categorize the cohort into two SCD patient samples. One sample was based on local clinical practice - MCI was excluded according to the Winblad criteria (this sample was termed SCDWinblad). The other sample was created via the retrospective application of the Jak/Bondi criteria for the exclusion of MCI (termed SCDJak/Bondi). Only patients aged ≥ 55 years at baseline with ≥ 12 months follow-up were considered for inclusion. The initial clinical/demographic characteristics of the samples were compared. Rates of incident dementia were calculated for each sample, and unadjusted and Mantel-Haenszel-adjusted incidence rate ratios were calculated to compare dementia incidence between the SCD samples. RESULTS: The Essex Memory Clinic database included 2,233 patients in total. The SCD and study eligibility criteria were used to select SCDWinblad (n = 86) and SCDJak/Bondi (n = 185) samples from the database. Median follow-up (3 years) did not differ between the two samples. The SCDJak/Bondi sample was significantly older than the SCDWinblad at first assessment (median age: 74 versus 70 years) and had poorer scores on tests of global cognition, immediate and delayed verbal recall, and category fluency. Following adjustment for age, the dementia incidence rate ratio [95% confidence interval] was 3.7 [1.5 to 9.3], indicating a significantly greater rate of progression to dementia in SCDJak/Bondi. CONCLUSIONS: This study highlights that the approach used to ascertain SCD has important implications for both SCD phenotypes and prognosis. This underscores the importance of how MCI is operationalized within SCD studies. More broadly, the findings add to a growing body of work indicating that objective cognition should not be overlooked in SCD, and offer a potential explanation for the heterogeneity across the SCD prognostic literature.