Objective: In the context of the positive impact of cognitive interventions for age-related memory concerns, clinicians are seeking information about variables that predict optimum client response. In this study of older adults, the aim was to investigate baseline predictors of gain in memory performance, i.e. prospective memory, following a memory intervention.Methods: One hundred and one healthy older adults (H0A) and 73 older people with amnestic mild cognitive impairment (aMCI) were evaluated at 6-months after participating in a 6-week memory group intervention (LaTCH). The outcome measure was a clinic-based prospective memory task. Baseline predictors included demographic variables (age, gender, education), baseline prospective memory, and cognitive resources (retrospective memory, executive function).Results: Thirty percent of the HOA and 16% of the aMCI cohorts demonstrated reliable training effects on prospective memory test performance at 6-month assessment. Through hierarchical regressions in the HOA cohort, executive function (working memory, attention set shifting) rather than retrospective memory was the best predictor of change in prospective memory. Moderated regression did not demonstrate any interactions between retrospective memory and executive function. For the memory impaired cohort (aMCI), better baseline retrospective memory predicted greater gain in prospective memory but only when executive function was also high.Conclusions: Memory groups can improve performance on clinic-based prospective memory tests in older people with concerns about memory performance, suggesting the value of further translation studies to demonstrate functional real-world gains and quality of life improvement after training. These interventions may be especially effective for those older people with better executive function (working memory, attention set shifting).