BACKGROUND accurate classification of older people into fallers and non-fallers is crucial for falls research, but largely dependent on the accuracy of fall reporting by the participants. OBJECTIVE to investigate the influence of memory in relation to fall reporting. METHODS five hundred community-dwelling adults aged 70-90 years. METHODS memory and executive functioning were assessed using the Rey Auditory Verbal Learning and Trail Making test, respectively. Fall risk was estimated using the physiological profile assessment (PPA). Falls were recorded prospectively for 12 months using monthly falls diaries and follow-up phone calls as required. RESULTS Spearman correlations showed that falls were significantly correlated to worse executive functioning, worse PPA scores and better memory. People with better memory had an increased risk of being classified as single fallers and multiple fallers, but not when reported injuries were included as part of the definition. CONCLUSIONS good memory appears to influence the recording of falls in community-dwelling older people and likely reflects a reporting bias. In research studies, there may be value in using a combination of injurious falls and multiple falls when classifying people into faller and non-faller groups.