Background: While medications may prolong life and prevent morbidity in older people, adverse effects of polypharmacy are increasingly recognised. As patients age and become frail, prescribing may be expected to focus more on symptom control and minimise potentially harmful preventive medication use that confer little benefit within a short lifespan. Whether prescribing practice shifts to one of symptom controls among the oldest old admitted to hospital remains unclear. Aim: To determine, in the oldest old inpatients, whether preventive versus symptom control medication prescribing was associated with age or level of frailty. Methods: Retrospective analysis of all patients aged ≥85 years referred for comprehensive geriatric assessment at a tertiary care hospital between May 2006 and December 2014 for whom all prescribed medications were documented. Medication use was assessed according to age group (85–89, 90–94, ≥95) and categories of frailty index calculated for patients based on 52 deficits (fitter, moderately frail, frail and severely frail). Results: Seven hundred and eighty-three inpatients were assessed of mean (SD) age 89.0 (3.4) and mean frailty index 0.45 (SD 0.14) with a median of eight co-morbidities (IQR 6–10) and who were prescribed a mean of 8.3 (SD 3.8) regular medications per day. Polypharmacy (5–9 medications per day) was observed in 406 patients (51.9%) and hyper-polypharmacy (≥10 medications per day) in 268 patients (34.2%). While there was a significant decrease in number of prescribed medications as age increased, there were no differences across age groups or frailty categories in proportions of medications used for prevention versus symptom control. Conclusion: Polypharmacy is prevalent in oldest old inpatients and prescribing patterns according to prevention versus symptom control appear unaffected by age and frailty status.