To determine whether the 'Timed Up and Go' (TUG) test is a useful test for predicting re-attendance at an ED, emergency hospital admission or death within 90 days in elderly patients discharged from the ED. This was a prospective blinded cohort study at a tertiary referral ED. Patients completed a TUG test during their Allied Health assessment prior to discharge from the department. After 90 days, patient ED attendances, emergency admissions to hospital or deaths were recorded and confirmed by phone. Data were analysed using logistic regression and reported as odds ratios (OR) or log-transformation and Pearson analysis. One hundred patients were enrolled: 78 (78%, 95% confidence interval [CI] 70-86%) patients remained event free, 22 (22%, 95% CI 14-30%) patients re-attended an ED and 15 (15%, 95% CI 8-22%) were admitted to hospital as an emergency admission. There was no significant difference between TUG test times and whether patients re-attended an ED (OR 1.0 [0.93-1.06] P = 0.9) or were admitted to hospital (OR 0.99 [0.91-1.07] P = 0.74). There was no significant correlation between a patient's TUG test time and the number of days to ED re-attendance (Pearson correlation coefficient 0.38 [-0.04 to 0.69] P = 0.08) or admission (Pearson correlation coefficient 0.32 [-0.23 to 0.71] P = 0.25). This study did not detect any predictive value of the TUG test for ED re-attendance or hospital admission within 90 days of discharge among aged ED patients.