Objectives: To use local ambulance service patient care records (PCRs) at an aggregate level to study the use of emergency medical services (EMS) by older people with dementia in two English counties. To understand how and where in the PCR dementia is recorded. To measure the proportion of patients aged 75 and over who had an emergency (999) ambulance response who have dementia recorded in the PCR. To carry out a descriptive analysis of any associations with age, gender, reason for the call, time of call, residential status or call outcome. Methods: Four days of PCRs from two counties (UK) for patients aged 75 and over were reviewed and non-patient-identifiable data extracted. Data for the total number of call-outs for those days were obtained from the computer-aided dispatch (CAD) system. Results: In 4 days' records for Cambridgeshire and Hertfordshire (2 304 records), over one third of call-outs (830) were to patients aged 75 and over. Data were obtained from 358 paper records. Dementia was recorded on 14.5% of records and another 7.0% recorded details suggesting dementia or cognitive impairment. Around 15% of call-outs to ≥75-year-olds were to care homes. Ambulance crews attended higher proportions of ‘older old’ people than the local population percentages of 85 to 89-year-olds and ≥90-year-olds. The most common reason (27.5%) for a call-out was a fall. Conclusions: This is the first paper to look in detail at the numbers of older people with dementia seen by emergency ambulance crews as documented in PCRs. It gives a benchmark for others looking at ambulance service data and highlights possibilities and pitfalls of using ambulance service PCR data.