Background: Emergency departments are experiencing crowding and delays in care, which are associated with poorer patient outcomes. Nurse staffing is a key determinant of patient throughput and outcomes, but much of the existing research relies on indicators such as average nurse-to-patient ratios that do not account for daily variation in patient acuity. Acuity-adjusted measures of staffing that reflect the balance between available nursing hours and patient care requirements may provide a more accurate understanding of how nurse staffing affects emergency department performance.Objective: To examine the association between nurse staffing, adjusted for daily case-mix acuity, and patient processes and outcomes. Design: Retrospective longitudinal study.Setting: Emergency department of a university hospital. Participants: Administrative data from 67,010 patients corresponding to 115,078 adult visits over 787 consecutive days.Methods: Routinely collected emergency department data from January 2018 to February 2020 were analysed, with each 24-hour period as the unit of analysis. Actual nursing hours were taken from rosters. Required nursing hours were estimated by linking each patient’s Manchester Triage System category to a pre-determined nursing care time. A staffing deficit was defined as actual minus required hours. Process measures were time to triage, and wait to be seen; outcome measures were patient experience time, and leaving without being seen. Models were run unadjusted and adjusted for potential confounders.Results: On average, 85% of visits were triaged as urgent or very urgent. On 88% of days, actual nursing hours were below the estimated requirement. Median times were 21 minutes to triage, 2.1 hours to be seen, and a median of 7.0 hours total patient experience time. Staffing deficits were associated with poorer patient processes and outcomes (p < 0.001). Each 5-hour deficit (versus no deficit) in nurse staffing was associated with a 21% increase in time to triage (RGM 1.21; 95% CI 1.19 to 1.23), a 12% increase in wait to be seen (RGM 1.12; 95% CI 1.09 to 1.14), a 7% increase in patient experience time (RGM 1.07; 95% CI 1.05 to 1.09), and 16% higher odds of a patient leaving without being seen (OR 1.16; 95% CI 1.11 to 1.22). Conclusions: Lower nurse staffing relative to patient acuity was associated with delays across all stages of care and more patients leaving without being seen. An acuity-adjusted staffing metric provides a more accurate relationship between staffing and patient need than simple nurse-to-patient ratios supporting its use in determining safe emergency department staffing.