Objective. To examine the implementation of estimated date of discharge (EDD) for planned admissions and admissions via the emergency department, to assess the variance between EDD and the actual date of discharge (ADD), and to explore the determinants of delayed discharge in a tertiary referral centre, Sydney, Australia. Methods. Primary data from a convenience sample of 1958 admissions for allocation of EDDs were linked with administrative data. The window for assigning EDDs for planned admissions was 24 h, for admissions via the emergency department it was 48 h. Logistic regression models were used to examine the key factors associated with an EDD being assigned within 24 h or 48 h of an admission. An ordinal logistic regression model was used to explore the determinants of delayed discharge. Results. Only 13.4% of planned admissions and 27.5% of admissions via the emergency department were allocated a timely EDD.Older patients, patients with significant burdens of chronic morbidity(OR = 0.903;P = 0.011); and patients from a non-English-speaking background (OR = 0.711; P = 0.059) were less likely to be assigned a timely EDD. The current Charlson Index score was a significant predictor of a positive variance between EDD and ADD. Conclusions. The prevalence of the timely assignment of an EDD was low and was lowest for planned admissions. The current Charlson Index score is an effective tool for identifying patients who are more likely to experience delayed discharge.